Re: [time-nuts] Cycling of Peltier junction
There is a reference to thermomechanical fatigue in Peltier devices with respect to the effect of peltier device ripple current on the longevity of the Peltier devices in the HP Journal article on an optical power meter. Bruce J. Forster wrote: Peltier devices have been used as temperature control elements for decades. I've never heard of fatigue failures, but, if I were designing a chamber as you suggest, I'd try to keep the temperature differential across the TE element under maybe 15 to 20F. The harder you push it, the greater the stress. Also, the heat pumping power falls dramatically as the delta-T increases and it's a situation of rapidly deminishing returns. I'd not worry much about ramping the drive. FWIW, -John === Does anybody know about using the same Peltier junction for both heating and cooling? I'm concerned about thermal/mechanical shock when changing the polarity back-n-forth between hot and cold. Maybe there needs to be a controlled ramp, if so then how do I figure out the rate? Why: I'm in the process of building a small environmental chamber for my home lab. The volume is ~30 liter, target temp range of 0C to 60C. For the cooling side I am using water circulation (radiator, pump, reservoir water block) and Peltier junctions. At first I was planning to have two separate systems, one for heating and one for cooling, but then I got to thinking that using just the water and Peltier could be used for both. I will be using a PID for temp control, and two TEC1-12726 Peltier Qcmax(w)= ~240 $B$(BT =0j Regards, Jerome --- This email message is for the sole use of the intended recipient(s) and may contain confidential information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. --- ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Freestanding mast
Magnus Danielson wrote: Hi Steve, On 09/05/2010 10:18 AM, Steve Rooke wrote: On 5 September 2010 04:42, Rob Kimberleyr...@timing-consultants.com wrote: Just a thought, as you are in southern hemisphere, wouldn't you see more birds facing North? Oops! I really meant North. Well spotted that man. My satellite azimuth/elevation chart looks quite typical to text-book style. My GPSDOs still seem to be recovering from the long power outage caused by the earthquake here early Saturday morning but the stats seem to be settling down again. My timing gear and antenna were unaffected but it sure moved some of the heavy HP instruments that I have piled up on my workbench and demolished my computer rack, but luckily everything seems to be working OK. The only thing that seems to be at fault is my broadband which is playing up now and I wonder if the telephone lines have been damaged in some way. I was about to ask how you New Zeeland time-nuts was doing and affected by the earthquake. Cheers, Magnus Along with other North Island dwelling TN's I didn't feel a thing. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Maser manual
Attila Kinali wrote: On Wed, 01 Sep 2010 20:08:13 + Poul-Henning Kampp...@phk.freebsd.dk wrote: Yes, but what is the issues relating to sapphire loading? What's the cost of the sapphire block and having it machined? It is a saphire tube, a readily available, if not exactly cheap, commodity. Why saphir? Aluminia (AlO2) seems to be used as well to load H maser cavities. Or is saphir in some way better? Attila Kinali Sapphire and ruby are slightly impure varieties of corundum the single crystal form of aluminium oxide. Sapphire and rubies just have different inpurities that impart colour to the gem. The microwave loss in single crystal alumina (sapphire, corundum) may be somewhat lower than for the polycrystalline form. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] homebrew H maser
jimlux wrote: Hal Murray wrote: jim...@earthlink.net said: The gas diffusing out through the drilled bolt.. sure it's drilled, but the conductance is so patheticaly low, you're literally waiting until the gas molecules happen to randomly bounce their wey up the hole. I've never worked with vacuum gear. I assume drilled bolt refers to a bolt through a drilled hole so there is some slop between the bolt and the hole. No... the bolt has a hole through it, to provide a gas path when you install it into a blind tapped hole. Otherwise, the trapped gas in the bottom of the hole slowly leaks out past the threads. Can I use vacuum grease as a seal around the bolt? Or does it outgas too much if you are going for seriously low pressures? For the most part, grease is more trouble than it's worth. Knife edge seals are where it's at. Can I use a soft(er) metal washer and mash it to a gas tight fit by tightening the bolt enough? Not exactly.. what you see is a knife edge cutting into a softer metal... mashing implies gas trapped between layers.. That kind of thing crops up in TWT manufacturing, where they stack all the parts of the gun or the collector... How low a pressure does a H maser need? Where is it relative to say fingerprints outgassing? That's a good question.. I don't know. http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2doc=GetTRDoc.pdfAD=ADA503712 http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2doc=GetTRDoc.pdfAD=ADA503712 Indicates that the operating pressure at the hydrogen dissociator is likely to be a few Torr or so. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Maser info
Higher operating temperatures force the use of nickel alloy to replace the silver palladium alloy traditionally used. At higher operating temperatures (40c and above) its not possible to turn off (without cooling it) the palladium leak. The Russian masers use nickel or nickel alloy instead of palladium or palladium silver. Bruce Corby Dawson wrote: John F., The Palladium valve is also known as a palladium leak or a palladium purifier. In the Maser the use is as the leak. It would also serve to purify the H2 BUT any other impurities lodge in the Palladium plug and can eventually cause it to fail. Early symptoms manifest as having to heat the plug to higher and higher temperatures to maintain the H2 flow. When the hydrogen bottle is changed you must perform a purge routine (see the manual) to allow any foreign gases to be removed. So for maximum life the Hydrogen should be as pure as possible. The resonator coil cannot be seen, I can see that it is a bit different than the manual shows. It was upgraded at some point. I can provide a picture of another masers coil. The receiving tank did not heat up all. Since I was going from a higher pressure tank to a mostly empty tank I don't think compression was involved Robert, You CANNOT use oil diffusion pumps, even for the rough pumping! (mechanical roughing pumps are also a no-no.) ANY contamination can seriously degrade the bulb coating. This can take quite a while to show up. Since tearing down the maser to replace the storage bulb is definitely NON-TRIVIAL. Using a turbo pump or vacsorbs are the only options. I use Vacsorbs as they are simple and quite a bit cheaper than the turbo. Bill, If your serious, the disassociator splits the hydrogen molecules H2 into atoms H to allow maser operation. I do have an old Interocitor screen I could mount on top of the Maser. It would look kinda neat! John M., The original oscillator was an upgrade and did not agree completely with the manuals schematic. I did get an updated schematic from the vendor and after much work and completely rebuilding it I still could not get it to work reliably. I decided to design my own using a low power oscillator to drive a power amplifier and impedance matching network. This has worked very well! Corby Dawson 1 Tip for Losing Weight Cut down 2 lbs per week by using this 1 weird old tip http://thirdpartyoffers.juno.com/TGL3141/4c7f52071849adf3adm04duc ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Maser info
Reference for palladium-silver leak difficulty at high temperature. http://tycho.usno.navy.mil/ptti/1988/Vol%2020_10.pdf Bruce Bruce Griffiths wrote: Higher operating temperatures force the use of nickel alloy to replace the silver palladium alloy traditionally used. At higher operating temperatures (40c and above) its not possible to turn off (without cooling it) the palladium leak. The Russian masers use nickel or nickel alloy instead of palladium or palladium silver. Bruce Corby Dawson wrote: John F., The Palladium valve is also known as a palladium leak or a palladium purifier. In the Maser the use is as the leak. It would also serve to purify the H2 BUT any other impurities lodge in the Palladium plug and can eventually cause it to fail. Early symptoms manifest as having to heat the plug to higher and higher temperatures to maintain the H2 flow. When the hydrogen bottle is changed you must perform a purge routine (see the manual) to allow any foreign gases to be removed. So for maximum life the Hydrogen should be as pure as possible. The resonator coil cannot be seen, I can see that it is a bit different than the manual shows. It was upgraded at some point. I can provide a picture of another masers coil. The receiving tank did not heat up all. Since I was going from a higher pressure tank to a mostly empty tank I don't think compression was involved Robert, You CANNOT use oil diffusion pumps, even for the rough pumping! (mechanical roughing pumps are also a no-no.) ANY contamination can seriously degrade the bulb coating. This can take quite a while to show up. Since tearing down the maser to replace the storage bulb is definitely NON-TRIVIAL. Using a turbo pump or vacsorbs are the only options. I use Vacsorbs as they are simple and quite a bit cheaper than the turbo. Bill, If your serious, the disassociator splits the hydrogen molecules H2 into atoms H to allow maser operation. I do have an old Interocitor screen I could mount on top of the Maser. It would look kinda neat! John M., The original oscillator was an upgrade and did not agree completely with the manuals schematic. I did get an updated schematic from the vendor and after much work and completely rebuilding it I still could not get it to work reliably. I decided to design my own using a low power oscillator to drive a power amplifier and impedance matching network. This has worked very well! Corby Dawson 1 Tip for Losing Weight Cut down 2 lbs per week by using this 1 weird old tip http://thirdpartyoffers.juno.com/TGL3141/4c7f52071849adf3adm04duc ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] homebrew H maser
Poul-Henning Kamp wrote: In message4c7f5918.7030...@xtra.co.nz, Bruce Griffiths writes: Indicates that the operating pressure at the hydrogen dissociator is likely to be a few Torr or so. The pressure is basically: As low as possible in order to minimize hydrogen collisions (other hydrogen, walls) as much as possible. i.e. the mean free path of the atomic hydrogen needs to be somewhat larger than the dimensions of the (fused silica) gas containment bulb. The mean free path will be comparable to the bulb dimensions at a pressure of around 1 ubar (100 uPa) or so. Since the Hydrogen atom bounces of the fluoropolymer coated walls thousands of times before phase coherence is lost the mean free path needs to be several thousand times the containment bulb dimensions to avoid degrading the maser performance. This requires a pressure of around 1 nanobar (100nPa) or below within the storage bulb.. The (gas) conductance of the exit aperture of the dissociator is selected to achieve the required atomic hydrogen flux of at most around 3E-5 liter-Torr/sec or so for a typical hydrogen dissociator pressure of 50Torr or so. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] homebrew H maser
Bruce Griffiths wrote: Poul-Henning Kamp wrote: In message4c7f5918.7030...@xtra.co.nz, Bruce Griffiths writes: Indicates that the operating pressure at the hydrogen dissociator is likely to be a few Torr or so. The pressure is basically: As low as possible in order to minimize hydrogen collisions (other hydrogen, walls) as much as possible. i.e. the mean free path of the atomic hydrogen needs to be somewhat larger than the dimensions of the (fused silica) gas containment bulb. The mean free path will be comparable to the bulb dimensions at a pressure of around 1 ubar (100 uPa) or so. Since the Hydrogen atom bounces of the fluoropolymer coated walls thousands of times before phase coherence is lost the mean free path needs to be several thousand times the containment bulb dimensions to avoid degrading the maser performance. This requires a pressure of around 1 nanobar (100nPa) or below within the storage bulb.. The (gas) conductance of the exit aperture of the dissociator is selected to achieve the required atomic hydrogen flux of at most around 3E-5 liter-Torr/sec or so for a typical hydrogen dissociator pressure of 50Torr or so. Bruce Oops!, the pressures given in Pa above are out a few orders of magnitude. Correct values are: The mean free path will be comparable to the bulb dimensions at a pressure of around 1 ubar (0.1Pa) or so. Since the Hydrogen atom bounces of the fluoropolymer coated walls thousands of times before phase coherence is lost the mean free path needs to be several thousand times the containment bulb dimensions to avoid degrading the maser performance. This requires a pressure of around 1 nanobar (100uPa) or below within the storage bulb.. The (gas) conductance of the exit aperture of the dissociator is selected to achieve the required atomic hydrogen flux of at most around 3E-5 liter-Torr/sec or so for a typical hydrogen dissociator pressure of 50Torr or so. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Maser manual
Magnus Danielson wrote: On 09/01/2010 09:39 PM, Poul-Henning Kamp wrote: In message24c547b54ea34a69bacc4f823bb40...@pc52, Tom Van Baak writes: I found the original copies of both EFOS manuals, along with a few photos. See: http://www.leapsecond.com/museum/efos/ Interesting. Page 4/3 in the service manual states: For the Hydrogen Maser, this unperturbed frequency is f(H) = 1 420 405 751.768 +/- 0.002 Hz In practice, this frequency is perturbed by interaction of the hydrogen atoms with the walls of the interaction volume container, doppler effects, interactions between the atoms themsel- ves, etc. The resulting frequency for the EFOS Maser is taken to be F(o) = 1 420 405 751.689 Hz I have no idea where the EFOS was produced, but somebody should try to calculate the relativistic correction for their height above the geoid, and see how much of the systematic 0.079Hz frequency difference that explains... Neuchatel, which still leaves a bit of unspecified height. However, this effect would be cancelled as their cesium clocks would be on the same height above the geoid (give or take a few meters). So, their indication is correct. The C-field also pulls the atoms of course, which they failed to point out in the cited text. If I were to build a maser myself, I would probably not attempt to copy the EFOS, as the large mechanical dimensions add significant cost in materials and machining. I would be much more tempted by a sapphire loaded cavity design like this one: http://www.nict.go.jp/publication/shuppan/kihou-journal/journal-vol50no1.2/0304.pdf) As that brings the mechanics inside the work envelope of main-stream CNC machines with the required tolerances. Yes, but what is the issues relating to sapphire loading? What's the cost of the sapphire block and having it machined? The tempco of the dielectric constant of sapphire is quite large so the cavity resonance tempco is somewhat larger than that of an unloaded copper or aluminium cavity. There is a NIST paper detailing a somewhat earlier attempt to use a dielectric cavity: http://tf.nist.gov/general/pdf/156.pdf Again the dielectric constant tempco is a significant issue. Cheers, Magnus Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Maser manual
An early analysis of a fused silica loaded cavity by Sigma Tau: http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA497003Location=U2doc=GetTRDoc.pdf http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA497003Location=U2doc=GetTRDoc.pdf Although the dielectric constant tempco and thermal expansion tempco of fused silica is low so is the dielectric constant so the reduction in cavity volume is relatively small. The reduced Q of a dielectric loaded cavity may also be an issue in the absence of cryogenic cavity cooling. Bruce Poul-Henning Kamp wrote: In message4c7eb534.2040...@xtra.co.nz, Bruce Griffiths writes: The tempco of the dielectric constant of sapphire is quite large so the cavity resonance tempco is somewhat larger than that of an unloaded copper or aluminium cavity. Yes, they write that cavity autotuning is a must. I still think that is a smaller problem than getting hold of and maching an unloaded cavity with the necessary shields. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] homebrew maser
PTFE wall storage bulb wall coatings haven't been used for some decades, FEP (or the Russian fluoropolymer ) is better in that a smoother coat is achievable see: http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2doc=GetTRDoc.pdfAD=ADA509340 http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2doc=GetTRDoc.pdfAD=ADA509340 A sual hexapole state selector is probably a little more effective than the cruder method used in the Russian masers. Bruce Mark Sims wrote: Same general idea, but an image intensifier plate would probably not work well. They are usually thinner and are cut at a bias so the electrons ricochet along its length. You might be able to mount one so that it cancels the bias angle. They are made by stretching a bundle of hollow glass tubes that have been filled with solid glass rods of a different composition. The original bundle can be very large (like over a meter) and is shrunk down to like 100 fibers per millimeter. It is then sliced and polished. Often the slices (or the pulled bundles) are joined into a bigger plate. Then the inner solid glass is dissolved out with a strong alkali. The hollow tubes are coated with a photoelectric material. The image from the tube is inverted using a twister... a coherent fiber optic rod that has a 180 degree twist. --- Do you know if the collimator is made from an uncoated microchannel plate? If so, an old, broken Gen II image intensifier might be a viable source. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] homebrew H maser
http://www.spectratime.com/product_downloads/PTTI_FCS_2005.pdf ewkeh...@aol.com wrote: Any links to reading material, would be nice to learn what they did to get a small package and how small is it? Bert Kehren In a message dated 8/29/2010 2:03:34 P.M. Eastern Daylight Time, p...@phk.freebsd.dk writes: In message3e227.34d2ee82.39abf...@aol.com, ewkeh...@aol.com writes: Do we know any thing about the Neuchatel design for Galileo? Bert Kehren There are plenty of papers about it. They started out with an active design, and got it inside spec (power/weight) but found that the performance was not worth the extra effort, so they switched to a passive design to further reduce weight/power. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] homebrew H maser
http://tycho.usno.navy.mil/ptti/ptti2002/paper14.pdf Bruce Griffiths wrote: http://www.spectratime.com/product_downloads/PTTI_FCS_2005.pdf ewkeh...@aol.com wrote: Any links to reading material, would be nice to learn what they did to get a small package and how small is it? Bert Kehren In a message dated 8/29/2010 2:03:34 P.M. Eastern Daylight Time, p...@phk.freebsd.dk writes: In message3e227.34d2ee82.39abf...@aol.com, ewkeh...@aol.com writes: Do we know any thing about the Neuchatel design for Galileo? Bert Kehren There are plenty of papers about it. They started out with an active design, and got it inside spec (power/weight) but found that the performance was not worth the extra effort, so they switched to a passive design to further reduce weight/power. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
Bob Camp wrote: Hi If you start with a mixer that runs 500 mV / radian (an RPD-1 at the typical 8 mV / degree + 10%) then -180 below that would be 0.5 nV. Since noise it coherent close in, the DSB to SSB process nets you 1 nV out when you have -180 dbc noise. With a capacitive IF port termination the mixer sensitivity increases somewhat. It increases more when using something like an HP10514 or 10534 than with an RPD-1. Such a termination isnt particularly useful for offsets much above 100kHz or so. If one terminates all mixer ports in 50 ohms as some insist is the best method, then the mixer phase sensitivity is much lower, in which case a somewhat lower noise preamp may be required. The posted plot does show (together with the noise plot for the HPS5.1 preamp) that the 2SK369 and the IF9030 have much lower flicker noise than the BF862. On Aug 22, 2010, at 10:51 AM, Bruce Griffiths wrote: So everything above (and an AD797 and likely an OP-37) will do better than 2 nV / Hz into 1K Hz. That would let you check oscillators in the below -170, but not below -180 range. You might or might not find such an oscillator in your junk box. They certainly do exist. On the plot above, both devices would let you do the same thing at 10 Hz. Now you are into the range of highly unlikely to find. At reasonable frequencies, -135 is doing pretty well at 10 Hz. Bragging rights start at about -140. Highly unlikely cuts in much past that 10 Hz offset. I'm not talking about a one of a kind piece of magic at NIST, but about what's in your junk box. The 2SK369 is still holding ok for -170 at 1 Hz. Even for one of a kind magic, that's pretty crazy. Unlikely to find (and really tough to measure) cuts in at about -120 at 1 Hz. At 0.1 Hz offset, you will need to run an instrument bandwidth below 0.01 Hz to get anything close to a good approximation to the noise. Most lab analyzers run 100X t to get enough data. That puts you out around 10,000 seconds for the run. That's a crazy long time. DC coupled offsets are likely to nuke that run just about every time. I'm by no means knocking the idea of having a good preamp. I'm only trying to point out that the numbers above are *way* past what a reasonable person might need to sort through their basement oscillator collection. 50 db is a lot of margin. Bob For an AC coupled sound card based spectrum analyser dc frequencies much below 2Hz or so are of little interest. Being able to calibrate the preamp + sound card frequency response using the thermal noise of a resistor is convenient. This is more difficult to achieve with a bipolar input stage as the amplifier input current noise is significant. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
Having a simple method of determining the preamp frequency response can be a useful diagnostic tool during development, particularly if one uses componets like super capacitors in the amplifier signal path. If one doesnt have a suitable offset source handy the mixer ports can be driven in near quadrature by the same signal and the dc output as a function of the relative phase shift between the 2 mixer inputs can be used. However neither method calibrates the phase noise frequency response of the system. Adding RF noise to one of the mixer inputs can be used to measure the frequency response of the system. If the RF noise source is uncalibrated but stable then it can be used to measure the relative frequency response. The results of a dc (or beat frequency) measurement of the gain can then be used to correct the results to obtain a calibrated frequency response. If one is using a capacitive or other non conventional mixer IF port termination, then knowing the relative frequency response can be vital. Bruce Bob Camp wrote: Hi I've always calibrated my phase noise setups to the phase slope of the mixer I'm using. It does involve switching gains, but it's a direct system calibration. Beat note is 360 degrees, so this chunk is x degrees and you got y mv over that chunk. Check the slope on the other side of the beat note to make sure it's the same. Do some math and you have a radian to volt transfer function. If you are sorting junk box OCXO's it's a pretty good way to do it. The only added steps are an independent measurement of the switched gain / gain flatness and a short circuit input check to estimate the noise floor. Both are an initial setup / one time only sort of thing with most amps. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Bruce Griffiths Sent: Tuesday, August 24, 2010 3:25 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Phase noise measurement (was - no subject) Bob Camp wrote: Hi CHOP Being able to calibrate the preamp + sound card frequency response using the thermal noise of a resistor is convenient. This is more difficult to achieve with a bipolar input stage as the amplifier input current noise is significant. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
The link isnt particularly useful as guests cant view the attachments and registration is disabled Bruce dk...@arcor.de wrote: Wenzel Audio Amp referred to in this email. Perfect! I drive with it a 3561A and a 7L5! Works for me. The only problem is getting any more 2SK369. Any recommendations? NXP BF862, available from digi-key. I have used it in a similar hookup with good success. Its virtue is the low noise voltage AND low input capacitance at the same time. You could deploy MANY of them in parallel until you get into the capacitance range of a single Interfet device. One heroic effort for audio is here: http://www.diy-audio-engineering.org/index.php?board=2.0 HPS5.1 I currently use 3 pairs of SSM2210 in front of a AD797. regards, Gerhard ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
http://www.synaesthesia.ca/LNschematics.html Is a better link, in that one can actually view the circuit schematics. There are a few simple refinements that will dramatically improve the low frequency PSRR of the single ended JFET circuits in the HPS5.1: 1) split the 3k3 resistor feeding the green LEDs into into 2 series 1k6 resistors and bypass the common node of these 2 resistors to ground. This low pass filters the noise current flowing in the LEDs due to power supply noise. 2) It would probably be even more effective if the base of the cascode transistor were driven by a voltage equal to the JFET source voltage plus about 3.7V. It should, for example, be possible to use a selected JFET to do this. 3) The output servo should drive the noninverting input of the opamp via a CBCS cascode (or equivalent) with a load resistor connected to the input stage positive supply rail. This should improve the PSRR dramatically. I use something similar in one of my low noise preamps albeit with a few LEDs in series with the resistor to provide most of the voltage drop as in my case the required voltage drop is reasonably predictable. This reduces the noise contribution from the servo integrator. Bruce Bruce Griffiths wrote: The link isnt particularly useful as guests cant view the attachments and registration is disabled Bruce dk...@arcor.de wrote: Wenzel Audio Amp referred to in this email. Perfect! I drive with it a 3561A and a 7L5! Works for me. The only problem is getting any more 2SK369. Any recommendations? NXP BF862, available from digi-key. I have used it in a similar hookup with good success. Its virtue is the low noise voltage AND low input capacitance at the same time. You could deploy MANY of them in parallel until you get into the capacitance range of a single Interfet device. One heroic effort for audio is here: http://www.diy-audio-engineering.org/index.php?board=2.0 HPS5.1 I currently use 3 pairs of SSM2210 in front of a AD797. regards, Gerhard ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
The noise measurements for the HPSs 5.1 preamp: http://www.synaesthesia.ca/LNmeasurements.html indicate that while the high frequency noise is about 2.2x lower than an optimised single ended 2SK369 preamp its flicker noise is far higher. If one uses 5 2SK369's connected in parallel the flicker noise should be even lower whilst the high frequency noise will be comparable/ If the feedback resistor values are reduced perhaps 3 @SK369BLs will suffice. Even lower flicker noise should be achievable if IF9030s are substituted for the 2SK369s. Bruce Bruce Griffiths wrote: http://www.synaesthesia.ca/LNschematics.html Is a better link, in that one can actually view the circuit schematics. There are a few simple refinements that will dramatically improve the low frequency PSRR of the single ended JFET circuits in the HPS5.1: 1) split the 3k3 resistor feeding the green LEDs into into 2 series 1k6 resistors and bypass the common node of these 2 resistors to ground. This low pass filters the noise current flowing in the LEDs due to power supply noise. 2) It would probably be even more effective if the base of the cascode transistor were driven by a voltage equal to the JFET source voltage plus about 3.7V. It should, for example, be possible to use a selected JFET to do this. 3) The output servo should drive the noninverting input of the opamp via a CBCS cascode (or equivalent) with a load resistor connected to the input stage positive supply rail. This should improve the PSRR dramatically. I use something similar in one of my low noise preamps albeit with a few LEDs in series with the resistor to provide most of the voltage drop as in my case the required voltage drop is reasonably predictable. This reduces the noise contribution from the servo integrator. Bruce Bruce Griffiths wrote: The link isnt particularly useful as guests cant view the attachments and registration is disabled Bruce dk...@arcor.de wrote: Wenzel Audio Amp referred to in this email. Perfect! I drive with it a 3561A and a 7L5! Works for me. The only problem is getting any more 2SK369. Any recommendations? NXP BF862, available from digi-key. I have used it in a similar hookup with good success. Its virtue is the low noise voltage AND low input capacitance at the same time. You could deploy MANY of them in parallel until you get into the capacitance range of a single Interfet device. One heroic effort for audio is here: http://www.diy-audio-engineering.org/index.php?board=2.0 HPS5.1 I currently use 3 pairs of SSM2210 in front of a AD797. regards, Gerhard ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
The Wenzel Audio amp is a little noisier than it need be and it has a poor PSRR, so that a very low noise power supply with low ripple is essential. Its not too hard to improve the PSRR and the input noise of such a current feedback amplifier. There are JFETS (IF9030) with similar noise floors and significantly lower flicker noise. However the minimum order from Interfet is about $250. For noise measurements on several JFETS (including the IF9030 and the 2SK369) see: /Ultra-Low-Noise High Input Impedance Amplifier for Low-Frequency Measurement Applications/ Felix A Levinson, IEEE Transactions on Circuits and Systems Vol 55 No 7, August 2008 pp1815-1821. Bruce ewkeh...@aol.com wrote: Hi I have the Hp phase noise system with the 35601A but use most the time the Wenzel Audio Amp referred to in this email. Perfect! I drive with it a 3561A and a 7L5! Works for me. The only problem is getting any more 2SK369. Any recommendations? Thanks Bert Kehren In a message dated 8/20/2010 6:54:05 P.M. Eastern Daylight Time, jmi...@pop.net writes: Would anyone else like to suggest a known good low phase noise buffer amplifier? Maybe something from a Fred Walls paper? You can always build HF isolation amps by rigging MMICs and attenuators together, but this will not reliably get you below -160 dBc/Hz. Bruce G. has given some good advice in this regard, with some circuit designs at http://www.ko4bb.com/~bruce/IsolationAmplifiers.html and elsewhere. I'm a fan of this version (also from Bruce): http://www.ke5fx.com/norton.htm This one has the advantage of simplicity. No weird parts, nothing that is likely to be out of production or hard to find, and dirt cheap. I've measured the broadband floor at near -170 dBc/Hz at 10 MHz, and its noise contribution at 100 Hz is below what the 3048A can see. These figures are adequate to measure any 10811-class OCXOs. A practical PN measurement system for 10811-class oscillators can be made by building two of those amplifiers and using them to drive pretty much any random double-balanced mixer found on eBay with +10 dBm LO specs or more. Both ports should be driven strongly to reject AM artifacts and avoid degrading the excellent noise floor offered by the amps. I'd hit the LO port with +10 to +12 dBm and the RF port with at least 0 dBm. Then, see the Wenzel app note here ( http://www.wenzel.com/documents/measuringphasenoise.htm ) to lock the two oscillators in quadrature and amplify the resulting baseband output. Any of several sound-card FFT programs can be used to generate an output graph, although if you want absolute calibration in dBc/Hz you need to be prepared to sweep the actual test setup from mixer output to FFT input to watch for various sources of flatness error. A combination of an AD7760-EVAL board and a Digilent Nexys2 can be used to construct an excellent baseband digitizer for the DC-1 MHz spectrum, but most of the time a good-quality 192-kHz sound card is fine for this sort of work. Most good crystal oscillators reach their broadband floor by 10 kHz, so there's no real need to go out to 1 MHz or more. -- john, KE5FX ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
ewkeh...@aol.com wrote: Thanks Bert In a message dated 8/21/2010 11:43:53 A.M. Eastern Daylight Time, dk...@arcor.de writes: Wenzel Audio Amp referred to in this email. Perfect! I drive with it a 3561A and a 7L5! Works for me. The only problem is getting any more 2SK369. Any recommendations? NXP BF862, available from digi-key. Don't these devices have relatively high flicker noise? I have used it in a similar hookup with good success. Its virtue is the low noise voltage AND low input capacitance at the same time. You could deploy MANY of them in parallel until you get into the capacitance range of a single Interfet device. The input capacitance is relatively noncritical in this application (phase noise measurement) since it is shunted by the much larger output capacitance of the low pass filter at the mixer IF port. One heroic effort for audio is here: http://www.diy-audio-engineering.org/index.php?board=2.0 HPS5.1 I currently use 3 pairs of SSM2210 in front of a AD797. regards, Gerhard Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase noise measurement (was - no subject)
Poul-Henning Kamp wrote: In message7c37.12cdef25.39a12...@aol.com, ewkeh...@aol.com writes: I am not seeing it, what should I use to measure it 3561 and 7 spec analyzer do not show it? It is probably the 3561 not the 70k that has the best chance. I am not aware of the precise characteristics of the noise, but it sounds somewhat like a boiling pot. I became aware of it first time when I ran a small class-A audio amplifier from a couple of, probably too, small VRLA's some years ago, just for the fun of it. With no input signal, the speakers would gurgle faintly and it took me some time to locate the source of the noise to the batteries. I would guess its amplitude correlates with the ratio of discharge current to plate area, since it is chemical/mechanical in nature. These days, I would build a super-cap battery instead if I needed a low-power PSU with low noise. Poul-Henning PS: also be aware that almost all VRLA's have a very nasty resonance frequency somewhere in the low MHz band. If you are after low noise, you should always decouple the battery good poly/plastic caps right at the terminals. NIST found that NiCd cells are very quiet at least for low load currents: http://tf.nist.gov/general/pdf/1133.pdf Thus batteries are useful as low noise voltage references or for providing the relatively low base current of a BJT in a low phase noise RF amplifier. Perhaps its the gelled electrolyte that is the source of the noise problem with VLRA batteries?? Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
J.D. Bakker wrote: At 23:49 +1200 14-08-2010, Bruce Griffiths wrote: J.D. Bakker wrote: At 19:01 +1200 14-08-2010, Bruce Griffiths wrote: J.D. Bakker wrote: However the ultimate test (other than breadboarding it) is to actually simulate the sampling process and look at the deviation of the sampled voltages from linearity. That's not a bad idea (the recent Simulation thread notwithstanding), I'll see if I can find the time to cobble something together. Suggested procedure: - Assume perfect ADC buffers (not unrealistic, some of the MCP6xxx parts have enough GBW and slew rate), and a 2V ADC reference. - Independent variables: * Number of ADC bits (8...12) * ADC input noise (model as AWGN, vary over 0...10LSB) * ADC aperture jitter (AWGN, 0...2ns) * ADC sample rate (1 or 2 MSPS) * Ramp rate (0.1/0.2/0.5/1V/us, to be varied by changing C1 and only C1) * For Circuit 3: Difference between ramp rates (0...10%, again through C1) - Have LTSpice generate a simulated ramp with enough time resolution (say 100ps), do linear interpolation if needed. - For each combination of independent variables: * Generate simulated ramp(s) * Run a realistic number of -100ns/0ns/+100ns calibrations (call it 100 runs) * Sweep the simulated offset from -500ns to 500ns in 1ns steps * For each simulated offset, do a few thousand measurement runs * Collect statistics - Plot RMS and 90%-limits for the recorded data. That should keep all eight cores busy for a day or so. Does that sound like a workable plan? If I feel up to it, I'll see if I can add the simple RC-filter to the mix, although I'm less confident about doing proper a priori weighed error curve fitting on that than on the simple linear ramps. Its probably more informative to look at the effect of the factors one at a time before doing the full blown simulation that includes all such factors. Start with an ideal current source to isolate the effects due to switch capacitance etc, then try a few real current sources. It should be a relatively simple case of a nonlinear least squares fit for the simple RC circuit. Simulating the effect of statistical calibration using a uniformly distributed set of time intervals. In this case the non linearity will be reflected in the non statistical variations in the histogram of frequencies for each ADC value. (I'd like to look at slower ramps/ADCs because the more I think about it the more I prefer the ADuC7024, with +/-1LSB INL @1MSPS over the +/-6LSB @2MSPS of the ATXMega. An added bonus of the ADuC is that it has a small on-chip PLA, which might allow me to do without a CPLD). In the case of the 3 diode TAC devised by Kasper Pedersen some compensation of diode capacitance modulation occurs if the diodes are matched. Hadn't seen that one yet. Looks interesting, but losing another two diode drops on top of the current source's compliance range may be a bit too tight for 3.3V operation. I've tried it in the simulator and on the bench, and it works quite well. I'll check again, but thats not consistent with what I found with a simulated 1mA current source. As I mentioned a few messages ago the ramp becomes much more linear (due to swamping of parasitics) when the current and the capacitor are increased tenfold. Tried it again on the bench with the values as in the attached sim file (SMD parts dead bug on a ground plane, with a FDV301N in series with a 10R resistor shorting the capacitor, and a resistor to set the current), and as far as I can eyeball it on my 100MHz scope it works as advertised. Not that a scope check is the last word in linearity, but at least there are no gross discrepancies with the simulator's results. Having said that, I'm open for other suggestions wrt the current source. JD Monte Carlo B. The deviations in the current visible in the simualtion are too small to be noticeable with a scope. Increasing the ramp capacitor value has little effect on nonlinearity due to the Early effect. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: leaching was, Alternative time interval interpolation technique
The TDEV plot for the OCXO in question which can be derived from its ADEV plot is perhaps a useful guide to the expected jitter when measuring a particular time interval. For long time intervals the phase noise much closer to the carrier than 5Hz will tend to dominate. Bruce Bob Camp wrote: Hi Single cycle jitter is a bit confusing when you talk about bandwidths of 5Hz to 20 MHz off a carrier. Since phase noise at 5 Hz does contribute to jitter over that bandwidth, an OCXO (with good phase noise close in) would be needed. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Attila Kinali Sent: Monday, August 16, 2010 11:53 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] OT: leaching was, Alternative time interval interpolation technique On Mon, 16 Aug 2010 07:58:01 +1200 Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Such low jitter oscillators are readily available. With some care (bandpass filtering) a cycle to cycle jitter of around 50fs or so is attainable with a Wenzel OCXO for example. Apropos Wenzel: Is there any distributor that sells them in single quantities? Or do i have to get them from Wenzel directly? And is there any price list available? However the time interval jitter degrades as the time interval increases. Achieving a cycle to cycle jitter of 1ps or so is relatively easy with a 10MHz or 100MHz OCXO having sufficiently low phase noise. Why an OXCO? AFAIK the temperature has only an effect on long term stability/drift, but doesn't affect short term effects (which cause the jitters). Or am i missing something? Attila Kinali ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: leaching was, Alternative time interval interpolation technique
Attila Kinali wrote: On Sun, 15 Aug 2010 14:59:43 +1200 Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Mea Culpa. Below is a link to the paper using SAW filters to achieve a sub ps time interval interpolator noise: http://cddis.gsfc.nasa.gov/lw16/docs/papers/las_4_Prochazka_p.pdf And the associated presentation: http://cddis.gsfc.nasa.gov/lw16/docs/presentations/las_4_Prochazka.pdf A quick skimming over the error analysis Panek made, suggests that the jitter of the clock source is the biggest contributor to measurement errors. But he never says how a clock source with such a low jitter is build. Although he references a few times a module build by Josef Kölbl of the Fachhochschule Deggendorf, there is no description available what kind of device that is. Does anyone have any pointers to recommended reading on the design of such low jitter oscillators? Attila Kinali Such low jitter oscillators are readily available. With some care (bandpass filtering) a cycle to cycle jitter of around 50fs or so is attainable with a Wenzel OCXO for example. However the time interval jitter degrades as the time interval increases. Achieving a cycle to cycle jitter of 1ps or so is relatively easy with a 10MHz or 100MHz OCXO having sufficiently low phase noise. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
J.D. Bakker wrote: At 08:30 +1200 14-08-2010, Bruce Griffiths wrote: J.D. Bakker wrote: 4) If the ADC(s) have a sufficiently wide full power bandwidth then one could just sample a pair of quadrature phased 250kHz sinewaves. As someone who's used to thinking in I/Q I must say I've always liked the elegance of this approach. Trouble is that I don't see a cheap/easy way to generate quadrature sines with low enough distortion/noise. Distortion isnt a great problem if its relatively small and stable as it can always be measured as part of the calibration process and its effect may then be compensated in software. Those are two large ifs, if you're going for a small/cheap implementation. Fully analog solutions can get messy, phase shifter hybrids are doable but would definitely need to be shielded from drafts and DDSes plus appropriate filters do the job but are still relatively expensive. And then there's motor/generators... Am I missing any method here? (on a semi-related note: some interesting work on ultralow distortion low frequency oscillators, employing quadrature signals to simplify the AGC, is being done here: http://www.prodigy-pro.com/diy/index.php?topic=26461.0) (I'm not sure why I'd want to use a synchronizer in this path. The way I see it the TAC operates as a linear phase detector, with the GPS PPS and the synthesized PPS as inputs. The microcontroller then applies the sawtooth correction to the measured time offset, and uses the result in a DPLL. There will, of course, be a synchronizer in the input line from the GPS PPS to the microcontroller, but that's only used for the FLL and for rough synchronization). Using a synchroniser allows the TAC output range to be combined with the coarse timestamp derived by sampling a counter clocked by the same clock as the synchroniser. I think we're looking at it from two different angles. What I read from your description is close to the traditional architecture such as used in the HP5335A, with a counter running at the system clock frequency for coarse measurement and a TAC to measure the remainder. What I'm planning to do is more akin a traditional PLL, with the TAC as the Phase Detector. For this to work I assume that a coarse FLL (using a counter) has already brought the oscillator within lock range. Is there any reason that method won't work, or can trivially be made to work better? Having a wide TAC range means that its resolution and noise depends critically on that of the ADC. Since some ADCs embeded within processor dont have true 12 bit performance this may limit the TAC resolution/noise to several nanosec rather than the desired 1ns or better. (The regenerated PPS output will indeed be derived from and synchronous with the VCXO/OCXO. It is also my intention to have the OCXO clock the microcontroller, either directly or through a prescaler, depending on whether the XO runs higher or lower than the max CPU clock). That ensures that all intermod products are harmonically or submultiples of the OCXO frequency. - Circuit 3 expands on this approach by having dual ramp generators, and having the ADC measure the voltage difference between the two. Not a good idea, as this requires accurate matching of the gains of the 2 TACs. Why? At that points they're not TACs yet, just ramp generators. Circuit 3 uses the difference between these ramps, and I believe it need not be constant. Assume there's a 1% difference in ramp rates; say C3 charges at 1V/us and C4 charges at 1.01V/us. [...] Since NP0/C0G caps are only available in 5% and 10% tolerance at best, matching gains to 1% will require using selected parts (adjust current source to compensate) or trimming. I picked the 1% figure out of the air, simply to have an example for the math. Even so, if required it would be easy to have the microcontroller trim the ramp rates through one of the on-chip DACs. However I don't believe that the ramp rate can't be dealt with in software through calibration. Software is probably best (if feasible) as this eliminates the parasitic capacitances and noise associated with trimmers and DACs. With a single ADC its not possible to correct the TAC nonlinearity since there are a wide range of possible output voltages from the first TAC for any given differential input to the ADC. Simulation indicates nonlinearity of the order of 1% or so in the ramp generator. This is largely due to the Early effect and semiconductor output capacitance modulation. Yeah, I noticed that. It helps a lot in the four-transistor mirror to have all transistors carry approximately equal amounts of current. Further linearization can be achieved by increasing the current, slowing the ramp rate and picking transistors with lower hFE for a given fT and/or higher VAF. An output resistance of up to 1M can be achieved, but it's the voltage-dependent capacitance that's hurting linearity. Lower hfe requires tighter
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
J.D. Bakker wrote: At 19:01 +1200 14-08-2010, Bruce Griffiths wrote: J.D. Bakker wrote: At 08:30 +1200 14-08-2010, Bruce Griffiths wrote: Using a synchroniser allows the TAC output range to be combined with the coarse timestamp derived by sampling a counter clocked by the same clock as the synchroniser. I think we're looking at it from two different angles. What I read from your description is close to the traditional architecture such as used in the HP5335A, with a counter running at the system clock frequency for coarse measurement and a TAC to measure the remainder. What I'm planning to do is more akin a traditional PLL, with the TAC as the Phase Detector. For this to work I assume that a coarse FLL (using a counter) has already brought the oscillator within lock range. Is there any reason that method won't work, or can trivially be made to work better? Having a wide TAC range means that its resolution and noise depends critically on that of the ADC. Since some ADCs embeded within processor dont have true 12 bit performance this may limit the TAC resolution/noise to several nanosec rather than the desired 1ns or better. No, the TAC range would only be wide enough to cover the expected spread of valid PPS pulses from the GPS (say +/-500ns...+/-1us). With some internal 12 bit ADCs that dont have true 12 bit you will barely achieve 1ns resolution with a 2us range. (I've thought a bit more about what you proposed, ie using the TAC to measure synchronizer delay. Problem is I'd like to use the timestamping counter that's internal to the CPU, and I see no way of getting at the output of its built-in synchronizer. This could of course be fixed by using an external timestamping counter/synchronizer, but that seems like a bit of a waste of resources). Surely you only need an external synchroniser (ie a dual D flipflop) clocked by the same clock (or at least one synchronous with it) as the internal counter? The internal synchroniser then only adds a fixed delay. (The regenerated PPS output will indeed be derived from and synchronous with the VCXO/OCXO. It is also my intention to have the OCXO clock the microcontroller, either directly or through a prescaler, depending on whether the XO runs higher or lower than the max CPU clock). That ensures that all intermod products are harmonically or submultiples of the OCXO frequency. Indeed. I prefer knowing where my birdies are (and preferably placing them where they do the least harm), rather than having them drift over time, frequency and temperature. The output compliance of your four transistor current mirror is limited to around 1.3V or so before the onset of saturation or gross nonlinearity. It's actually better than that, from what I can see from simulations and measurements. If the transistor currents are close to equal and the ramp rate isn't too high, output current stays within 1% up to ~1V, and the mirror saturates at 0.6-0.7V. This is with common small-signal transistors with an fT of a few hundred MHz. Really? There are 2xVbe + 1x diode drop to subtract from 3.3V ie somewhere from 1.8V -2.4V leaving a ramp amplitude of 1.5V to 1.1V depending on temperature and transistor current. That's what I thought when I first saw it and started counting junctions, but it's actually quite a bit better than that as the cross-coupling of the transistors steers current from saturating transistors into the bases of the opposing CE transistor. I found it in Barrie Gilbert's chapter on Bipolar Current Mirrors in the book Analogue IC Design: the current-mode approach; Google Books has a preview of much of this chapter. Simulation appears to indicate otherwise, distortion starts to rise as one of the mirror transistors nears saturation. One way to look at this is to look at variations in ramp charging current. However the ultimate test (other than breadboarding it) is to actually simulate the sampling process and look at the deviation of the sampled voltages from linearity. In the case of the 3 diode TAC devised by Kasper Pedersen some compensation of diode capacitance modulation occurs if the diodes are matched. I've tried it in the simulator and on the bench, and it works quite well. If you want to test it I suggest increasing the current source to 10mA, the cap to 10nF and starting with 150R for R1/R2 plus 10R emitter resistors for the CE transistors. I've tested it with the common European BC5xx/BC8xx-types, but LTSpice seems to like it with 2N3906s too. In that configuration, the ramp stays within +/-150uV of a linear approximation over a ramp range between 0 and 2V when ramping at 1V/us, which corresponds to +/-0.6LSB for a 12-bit ADC. I'll check again, but thats not consistent with what I found with a simulated 1mA current source. The capacitor charging current started to deviate significantly as saturation was approached. I also simulated other current sources with higher
[time-nuts] Alternative time interval interpolation technique
A method that measures the phase of a damped LC circuit oscillatory transient triggered by the event to be timestamped: http://risorse.dei.polimi.it/digital/products/2010/High frequency,high time resolution time-to-digital converter employing passive resonating circuits.pdf http://risorse.dei.polimi.it/digital/products/2010/High%20frequency,high%20time%20resolution%20time-to-digital%20converter%20employing%20passive%20resonating%20circuits.pdf A dual of the circuit is readily devised using a CMOS gate plus an open drain (or equivalent) gate output for damping/quenching. However the ADC employed needs to be able to capture a sample burst at a relatively high sample rate. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: leaching was, Alternative time interval interpolation technique
Mea Culpa. Below is a link to the paper using SAW filters to achieve a sub ps time interval interpolator noise: http://cddis.gsfc.nasa.gov/lw16/docs/papers/las_4_Prochazka_p.pdf And the associated presentation: http://cddis.gsfc.nasa.gov/lw16/docs/presentations/las_4_Prochazka.pdf Bruce Stanley Reynolds wrote: oh well no credit for me, but what happened to the missing space when so many other spaces made it thru as %20 ? - Original Message From: Hal Murrayhmur...@megapathdsl.net To: Discussion of precise time and frequency measurementtime-nuts@febo.com Sent: Sat, August 14, 2010 8:27:09 PM Subject: Re: [time-nuts] OT: leaching was, Alternative time interval interpolation technique stanley_reyno...@yahoo.com said: When they receive the request for the pdf they check to see what page referred the request if it wasn't their site then they assume some other web site leaching bandwidth. This other site pretends to serve the file but in fact it is still served by them. This pretend site doesn't pay for the bandwidth to serve the files, win for them lose for the unprotected server. Nice try, but that's not the problem this time. From the original message: bruce.griffi...@xtra.co.nz said: http://risorse.dei.polimi.it/digital/products/2010/High frequency,high time resolution time-to-digital converter employing passive resonating circuits.pdf http://risorse.dei.polimi.it/digital/products/2010/High%20frequency,high%20t ime%20resolution%20time-to-digital%20converter%20employing%20passive%20resona ting%20circuits.pdf The URL overflows a line and contains spaces. The second copy inside has %20 where the spaces go. You are supposed to remove the line breaks and put it back together. The problem is that there is a missing space between High frequency, and high time. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] PicTic Data
An ACAM GP2 evaluation board is available here: http://shop.omegacs.net/ However an ACAM GP1 would probably be a better fit in a DMTD with a beat frequency of 10Hz or more as the GP1 has a measurement range of 200ms. Bruce Stanley Reynolds wrote: Thanks for looking at my data that was what I was fishing for all along :-) I was looking at the article: A Small Dual Mixer Time Difference (DMTD) Clock Measuring System W.J. Riley Richard posted And thinking it would be nice to do the DSS and Mixer boards to go with the pictic II. Or changing the Pictic II to use the Acam TDC GP2 Time to Digital Converter; 2 channel w/65 ps resolution now under $30. as Bruce suggested a while back. My wants sure exceed my cans ;-) Stanley - Original Message From: John Milesjmi...@pop.net To: Discussion of precise time and frequency measurementtime-nuts@febo.com Sent: Sat, August 14, 2010 11:25:44 PM Subject: Re: [time-nuts] PicTic Data John glad you are getting good results and have something to compare to. Back to me who doesn't have any knowns but lots of guessing. Attached is a run with a box cover over the pictic, run is shorter ~ 800 seconds but the box does look like it helps. I need to do a lot more testing but sometimes I just get excited :-) I imported your .txt file alongside the traces I captured. Assuming it was taken with 1 Hz on both START and STOP, it looks like the attached. You're getting the exact sort of results that I see if I feed both the START and STOP inputs at 1 Hz. My guess is that the onboard oscillator limits the performance in that case, since it has a lot of time to drift during the measurement if the two pulses occur close to 1 second apart. Even the 5370B looks much worse if driven with 1 Hz on both inputs than it does with 1 Hz at START and 10 MHz at STOP. So I think you're basically up and running OK. When I get around to trying a better clock, I'll also go back and see if the 1-pps x2 performance improves. It would be great if the next spin of the board could include sine-to-CMOS shapers for the input channels as well as an external clock input, for people who are working directly with RF signals as opposed to 1-pps. -- john, KE5FX - Original Message From: John Milesjmi...@pop.net To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Sat, August 14, 2010 10:19:46 PM Subject: Re: [time-nuts] PicTic Data A few preliminary measurements here (I'm working on getting some software support together): http://www.ke5fx.com/pictic.htm -- john, KE5FX -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com]on Behalf Of Stanley Reynolds Sent: Saturday, August 14, 2010 7:12 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] PicTic Data My guess as to what the data may indicate is performance of the 10 Mhz 20PPM PICTIC internal oscillator, need to repeat test with precision 10Mhz and auto calibrate off. Fatness of the line/width maybe PICTIC error. Note graph seems to show me leaving the room and returning via the outside door. Not sure what the ~100 sec oscillations are, need to check a/c cycle time. Stanley ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] PicTic Data
Bruce Griffiths wrote: An ACAM GP2 evaluation board is available here: http://shop.omegacs.net/ Ignore this link as I omitted to read the fine print However an ACAM GP1 would probably be a better fit in a DMTD with a beat frequency of 10Hz or more as the GP1 has a measurement range of 200ms. Bruce Stanley Reynolds wrote: Thanks for looking at my data that was what I was fishing for all along :-) I was looking at the article: A Small Dual Mixer Time Difference (DMTD) Clock Measuring System W.J. Riley Richard posted And thinking it would be nice to do the DSS and Mixer boards to go with the pictic II. Or changing the Pictic II to use the Acam TDC GP2 Time to Digital Converter; 2 channel w/65 ps resolution now under $30. as Bruce suggested a while back. My wants sure exceed my cans ;-) Stanley - Original Message From: John Milesjmi...@pop.net To: Discussion of precise time and frequency measurementtime-nuts@febo.com Sent: Sat, August 14, 2010 11:25:44 PM Subject: Re: [time-nuts] PicTic Data John glad you are getting good results and have something to compare to. Back to me who doesn't have any knowns but lots of guessing. Attached is a run with a box cover over the pictic, run is shorter ~ 800 seconds but the box does look like it helps. I need to do a lot more testing but sometimes I just get excited :-) I imported your .txt file alongside the traces I captured. Assuming it was taken with 1 Hz on both START and STOP, it looks like the attached. You're getting the exact sort of results that I see if I feed both the START and STOP inputs at 1 Hz. My guess is that the onboard oscillator limits the performance in that case, since it has a lot of time to drift during the measurement if the two pulses occur close to 1 second apart. Even the 5370B looks much worse if driven with 1 Hz on both inputs than it does with 1 Hz at START and 10 MHz at STOP. So I think you're basically up and running OK. When I get around to trying a better clock, I'll also go back and see if the 1-pps x2 performance improves. It would be great if the next spin of the board could include sine-to-CMOS shapers for the input channels as well as an external clock input, for people who are working directly with RF signals as opposed to 1-pps. -- john, KE5FX - Original Message From: John Milesjmi...@pop.net To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Sat, August 14, 2010 10:19:46 PM Subject: Re: [time-nuts] PicTic Data A few preliminary measurements here (I'm working on getting some software support together): http://www.ke5fx.com/pictic.htm -- john, KE5FX -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com]on Behalf Of Stanley Reynolds Sent: Saturday, August 14, 2010 7:12 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] PicTic Data My guess as to what the data may indicate is performance of the 10 Mhz 20PPM PICTIC internal oscillator, need to repeat test with precision 10Mhz and auto calibrate off. Fatness of the line/width maybe PICTIC error. Note graph seems to show me leaving the room and returning via the outside door. Not sure what the ~100 sec oscillations are, need to check a/c cycle time. Stanley ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
Yet another option is to sample the output of a simple 1us time constant RC low pass filter and fit an exponential to the sampled data and calculate the threshold crossing from this. If the aberrations are sufficiently low over the range of time intervals measured (0.5us to 1us with a conventional 2 stage synchroniser clocked at 2MHz) this would be the simplest solution in that it only requires a single resistor and a single capacitor. Bruce Bruce Griffiths wrote: Yes, with a 2MSPS ADC and 1-2us transition times one gets 2-4 samples during the transition. Worst case with a 1us filter (10%-90%) output transition time there may be one sample at the midpoint and samples close to the 10% and 90% amplitude points. 2us transition times are probably close to optimum. In the latter case the effective time stamp resolution (with a true 12 bit ADC) will be around 0.5ns. Ideally a gaussian impulse response filter should be used. However if the input transitions are sufficiently (to allow the filter transients to settle) far apart almost any reasonable (without excessive overshoot) could be used. The minicircuits LPF_BOR3+ low pass filter appears almost good enough. Bruce Bob Camp wrote: Hi Would't you want 2 or more samples during the transition? Bob On Aug 12, 2010, at 8:25 PM, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Another method is to attenuate (to within the ADC input range) the PPS signal to be timestamped, low pass filter it and capture a 2MSPS sample burst centred around the low pass filter output transition midpoint. You can then use WKS interpolation to time stamp the transition midpoint (when it crosses a threshold halfway between the initial and final values of the low pass filter output). The low pass filter (preferably an LC filter) delay is easily calibrated by timestamping an internally generated signal initiated on a known ADC sampling clock edge. No (external) current sources, reset switches etc are required. With a 2MSPS sample rate a low pass filter output transition time of 1-2us should suffice (provided the ADC has a sufficiently large large signal bandwidth). Bruce Bruce Griffiths wrote: Some options: 1) Use a 74AHC05 for Q1 and Q2. 2) Switch the current source at the emitter node and only turn on the current source when charging the capacitor. This will increase the available TAC output voltage range and/or improve the linearity by eliminating the diode. However the capacitor discharge switch should be turned off before charging the capacitor. A stable fixed delay of a few (10ns??) before switching on the current source is required. 3) Replace the current source with a resistor. The resultant nonlinearity is well defined and software correction should be relatively easy. 4) If the ADC(s) have a sufficiently wide full power bandwidth then one could just sample a pair of quadrature phased 250kHz sinewaves. Extend the range by sampling (synchronise the input sampling edge to the counter clock first) a counter clocked at 250KHz. Initiate the sampling with the signal edge to be time stamped. If the GPSDO is used to clock the microprocessor, counters and produce the quadrature sinewave outputs then only a single TDC (time to digital converter) is required. Measuring negative time intervals should not be necessary as the TAC (or other TDC) should be used merely to measure the delay of a synchroniser the output of which is used to synchronously sample a counter clocked with the same clock as the synchroniser. J.D. Bakker wrote: Hello all, I'm working on Yet Another DIY GPSDO, and one of the issues I've been looking into is a TAC/TDC to do sawtooth correction on the measurement of the GPS PPS signal. I'd like to stick with a 3.3V supply for most of the circuit, and several of the TAC designs that have been discussed here in the past run into trouble at such low voltages (mostly through VBE drops). To start with the context: I'm planning to use a microcontroller with a built-in dual 12-bit 2MSPS ADC. I'd like to not use anything that's not available at Digi-Key or Mouser, and keep the SMD pitch=0.8mm (with a possible exception for dual transistors in SOT-23-6). That way the design shouldn't be too hard for others to replicate. I'm aiming for a TAC accuracy of 1ns, allowing for one or a few calibrations between PPS pulses. Minimum full-scale range should be +/- a few hundred ns, to allow for outliers. (The plan is to have an initial FLL for coarse locking, and have the PLL kick in after that). I'm penciling in an ADC reference voltage of 2V, as that's commonly available and leaves enough headroom to use the current sources in their most linear range. I've attached a diagram that reflects a few of my current thoughts. - Circuit 1 is the traditional TAC. Before the start of the cycle Q2 conducts, discharging C1 and shunting I1's current to ground. At this point the ADC can measure the voltage drop
Re: [time-nuts] one-off PC board
http://www.cordellaudio.com/papers/thd_analyzer.pdf Lester Veenstra wrote: Dick: I, for one, would be interested in knowing more about Bob Cordell's state-variable low-distortion oscillator. Do you soft copy details or a pointer to a source? Thanks, 73 Les Lester B Veenstra MØYCM K1YCM les...@veenstras.com m0...@veenstras.com k1...@veenstras.com US Postal Address: PSC 45 Box 781 APO AE 09468 USA UK Postal Address: Dawn Cottage Norwood, Harrogate HG3 1SD, UK ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
Some options: 1) Use a 74AHC05 for Q1 and Q2. 2) Switch the current source at the emitter node and only turn on the current source when charging the capacitor. This will increase the available TAC output voltage range and/or improve the linearity by eliminating the diode. However the capacitor discharge switch should be turned off before charging the capacitor. A stable fixed delay of a few (10ns??) before switching on the current source is required. 3) Replace the current source with a resistor. The resultant nonlinearity is well defined and software correction should be relatively easy. 4) If the ADC(s) have a sufficiently wide full power bandwidth then one could just sample a pair of quadrature phased 250kHz sinewaves. Extend the range by sampling (synchronise the input sampling edge to the counter clock first) a counter clocked at 250KHz. Initiate the sampling with the signal edge to be time stamped. If the GPSDO is used to clock the microprocessor, counters and produce the quadrature sinewave outputs then only a single TDC (time to digital converter) is required. Measuring negative time intervals should not be necessary as the TAC (or other TDC) should be used merely to measure the delay of a synchroniser the output of which is used to synchronously sample a counter clocked with the same clock as the synchroniser. J.D. Bakker wrote: Hello all, I'm working on Yet Another DIY GPSDO, and one of the issues I've been looking into is a TAC/TDC to do sawtooth correction on the measurement of the GPS PPS signal. I'd like to stick with a 3.3V supply for most of the circuit, and several of the TAC designs that have been discussed here in the past run into trouble at such low voltages (mostly through VBE drops). To start with the context: I'm planning to use a microcontroller with a built-in dual 12-bit 2MSPS ADC. I'd like to not use anything that's not available at Digi-Key or Mouser, and keep the SMD pitch =0.8mm (with a possible exception for dual transistors in SOT-23-6). That way the design shouldn't be too hard for others to replicate. I'm aiming for a TAC accuracy of 1ns, allowing for one or a few calibrations between PPS pulses. Minimum full-scale range should be +/- a few hundred ns, to allow for outliers. (The plan is to have an initial FLL for coarse locking, and have the PLL kick in after that). I'm penciling in an ADC reference voltage of 2V, as that's commonly available and leaves enough headroom to use the current sources in their most linear range. I've attached a diagram that reflects a few of my current thoughts. - Circuit 1 is the traditional TAC. Before the start of the cycle Q2 conducts, discharging C1 and shunting I1's current to ground. At this point the ADC can measure the voltage drop across C1/Q2 to eliminate that offset. Taking nSTART low puts Q2 into high-impedance, and I1 charges C1 through D1 until STOP is raised causing Q1 to shunt I1's current to ground. At this point the ADC samples the voltage across C1, which is proportional to the time between START and STOP (modulo offset and nonlinearities). This circuit is well known to work (although it is more common to use Q1 for both START and STOP and to limit Q2 to ramp discharge duties). Downsides are that negative time offsets cannot be measured directly, and the constant output voltage offers little room for increased precision through sample averaging, unless the ADC's input noise is large compared to its LSB size. For the same reason there is no easy way to reduce the effects of ADC INL/DNL. - Circuit 2 works in a similar way, except that the ramp isn't terminated by a STOP signal but is allowed to run freely until I2 saturates. The ADC is set to sample continuously, taking multiple samples of the ramp, and the microcontroller interpolates the resulting values to determine the elapsed time between an internal time reference point and the START signal. This circuit is fairly simple, and has the advantage that there is no hard limit to its range. Curve-fitting the sampled values increases precision and reduces the effects of INL/DNL. On the other hand, ADC aperture jitter and offset have a direct impact on resolution. - Circuit 3 expands on this approach by having dual ramp generators, and having the ADC measure the voltage difference between the two. Not a good idea, as this requires accurate matching of the gains of the 2 TACs. Its better to sample each TAC output individually as this allows software correction for gain mismatch (and nonlinearity) before subtraction. Software correction is better than using trimpots or similar as the parasitics etc associated with trimpots are eliminated. This approach is the only one of the three that can directly measure negative time offsets, allowing a regenerated pulse to be directly compared with the GPS' PPS. A small difference in ramp rates, unavoidable in practice, actually helps to average out
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
Another method is to attenuate (to within the ADC input range) the PPS signal to be timestamped, low pass filter it and capture a 2MSPS sample burst centred around the low pass filter output transition midpoint. You can then use WKS interpolation to time stamp the transition midpoint (when it crosses a threshold halfway between the initial and final values of the low pass filter output). The low pass filter (preferably an LC filter) delay is easily calibrated by timestamping an internally generated signal initiated on a known ADC sampling clock edge. No (external) current sources, reset switches etc are required. With a 2MSPS sample rate a low pass filter output transition time of 1-2us should suffice (provided the ADC has a sufficiently large large signal bandwidth). Bruce Bruce Griffiths wrote: Some options: 1) Use a 74AHC05 for Q1 and Q2. 2) Switch the current source at the emitter node and only turn on the current source when charging the capacitor. This will increase the available TAC output voltage range and/or improve the linearity by eliminating the diode. However the capacitor discharge switch should be turned off before charging the capacitor. A stable fixed delay of a few (10ns??) before switching on the current source is required. 3) Replace the current source with a resistor. The resultant nonlinearity is well defined and software correction should be relatively easy. 4) If the ADC(s) have a sufficiently wide full power bandwidth then one could just sample a pair of quadrature phased 250kHz sinewaves. Extend the range by sampling (synchronise the input sampling edge to the counter clock first) a counter clocked at 250KHz. Initiate the sampling with the signal edge to be time stamped. If the GPSDO is used to clock the microprocessor, counters and produce the quadrature sinewave outputs then only a single TDC (time to digital converter) is required. Measuring negative time intervals should not be necessary as the TAC (or other TDC) should be used merely to measure the delay of a synchroniser the output of which is used to synchronously sample a counter clocked with the same clock as the synchroniser. J.D. Bakker wrote: Hello all, I'm working on Yet Another DIY GPSDO, and one of the issues I've been looking into is a TAC/TDC to do sawtooth correction on the measurement of the GPS PPS signal. I'd like to stick with a 3.3V supply for most of the circuit, and several of the TAC designs that have been discussed here in the past run into trouble at such low voltages (mostly through VBE drops). To start with the context: I'm planning to use a microcontroller with a built-in dual 12-bit 2MSPS ADC. I'd like to not use anything that's not available at Digi-Key or Mouser, and keep the SMD pitch =0.8mm (with a possible exception for dual transistors in SOT-23-6). That way the design shouldn't be too hard for others to replicate. I'm aiming for a TAC accuracy of 1ns, allowing for one or a few calibrations between PPS pulses. Minimum full-scale range should be +/- a few hundred ns, to allow for outliers. (The plan is to have an initial FLL for coarse locking, and have the PLL kick in after that). I'm penciling in an ADC reference voltage of 2V, as that's commonly available and leaves enough headroom to use the current sources in their most linear range. I've attached a diagram that reflects a few of my current thoughts. - Circuit 1 is the traditional TAC. Before the start of the cycle Q2 conducts, discharging C1 and shunting I1's current to ground. At this point the ADC can measure the voltage drop across C1/Q2 to eliminate that offset. Taking nSTART low puts Q2 into high-impedance, and I1 charges C1 through D1 until STOP is raised causing Q1 to shunt I1's current to ground. At this point the ADC samples the voltage across C1, which is proportional to the time between START and STOP (modulo offset and nonlinearities). This circuit is well known to work (although it is more common to use Q1 for both START and STOP and to limit Q2 to ramp discharge duties). Downsides are that negative time offsets cannot be measured directly, and the constant output voltage offers little room for increased precision through sample averaging, unless the ADC's input noise is large compared to its LSB size. For the same reason there is no easy way to reduce the effects of ADC INL/DNL. - Circuit 2 works in a similar way, except that the ramp isn't terminated by a STOP signal but is allowed to run freely until I2 saturates. The ADC is set to sample continuously, taking multiple samples of the ramp, and the microcontroller interpolates the resulting values to determine the elapsed time between an internal time reference point and the START signal. This circuit is fairly simple, and has the advantage that there is no hard limit to its range. Curve-fitting the sampled values increases precision and reduces the effects of INL/DNL. On the other hand
Re: [time-nuts] On low-voltage TAC/TDCs for a GPSDO
Yes, with a 2MSPS ADC and 1-2us transition times one gets 2-4 samples during the transition. Worst case with a 1us filter (10%-90%) output transition time there may be one sample at the midpoint and samples close to the 10% and 90% amplitude points. 2us transition times are probably close to optimum. In the latter case the effective time stamp resolution (with a true 12 bit ADC) will be around 0.5ns. Ideally a gaussian impulse response filter should be used. However if the input transitions are sufficiently (to allow the filter transients to settle) far apart almost any reasonable (without excessive overshoot) could be used. The minicircuits LPF_BOR3+ low pass filter appears almost good enough. Bruce Bob Camp wrote: Hi Would't you want 2 or more samples during the transition? Bob On Aug 12, 2010, at 8:25 PM, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Another method is to attenuate (to within the ADC input range) the PPS signal to be timestamped, low pass filter it and capture a 2MSPS sample burst centred around the low pass filter output transition midpoint. You can then use WKS interpolation to time stamp the transition midpoint (when it crosses a threshold halfway between the initial and final values of the low pass filter output). The low pass filter (preferably an LC filter) delay is easily calibrated by timestamping an internally generated signal initiated on a known ADC sampling clock edge. No (external) current sources, reset switches etc are required. With a 2MSPS sample rate a low pass filter output transition time of 1-2us should suffice (provided the ADC has a sufficiently large large signal bandwidth). Bruce Bruce Griffiths wrote: Some options: 1) Use a 74AHC05 for Q1 and Q2. 2) Switch the current source at the emitter node and only turn on the current source when charging the capacitor. This will increase the available TAC output voltage range and/or improve the linearity by eliminating the diode. However the capacitor discharge switch should be turned off before charging the capacitor. A stable fixed delay of a few (10ns??) before switching on the current source is required. 3) Replace the current source with a resistor. The resultant nonlinearity is well defined and software correction should be relatively easy. 4) If the ADC(s) have a sufficiently wide full power bandwidth then one could just sample a pair of quadrature phased 250kHz sinewaves. Extend the range by sampling (synchronise the input sampling edge to the counter clock first) a counter clocked at 250KHz. Initiate the sampling with the signal edge to be time stamped. If the GPSDO is used to clock the microprocessor, counters and produce the quadrature sinewave outputs then only a single TDC (time to digital converter) is required. Measuring negative time intervals should not be necessary as the TAC (or other TDC) should be used merely to measure the delay of a synchroniser the output of which is used to synchronously sample a counter clocked with the same clock as the synchroniser. J.D. Bakker wrote: Hello all, I'm working on Yet Another DIY GPSDO, and one of the issues I've been looking into is a TAC/TDC to do sawtooth correction on the measurement of the GPS PPS signal. I'd like to stick with a 3.3V supply for most of the circuit, and several of the TAC designs that have been discussed here in the past run into trouble at such low voltages (mostly through VBE drops). To start with the context: I'm planning to use a microcontroller with a built-in dual 12-bit 2MSPS ADC. I'd like to not use anything that's not available at Digi-Key or Mouser, and keep the SMD pitch=0.8mm (with a possible exception for dual transistors in SOT-23-6). That way the design shouldn't be too hard for others to replicate. I'm aiming for a TAC accuracy of 1ns, allowing for one or a few calibrations between PPS pulses. Minimum full-scale range should be +/- a few hundred ns, to allow for outliers. (The plan is to have an initial FLL for coarse locking, and have the PLL kick in after that). I'm penciling in an ADC reference voltage of 2V, as that's commonly available and leaves enough headroom to use the current sources in their most linear range. I've attached a diagram that reflects a few of my current thoughts. - Circuit 1 is the traditional TAC. Before the start of the cycle Q2 conducts, discharging C1 and shunting I1's current to ground. At this point the ADC can measure the voltage drop across C1/Q2 to eliminate that offset. Taking nSTART low puts Q2 into high-impedance, and I1 charges C1 through D1 until STOP is raised causing Q1 to shunt I1's current to ground. At this point the ADC samples the voltage across C1, which is proportional to the time between START and STOP (modulo offset and nonlinearities). This circuit is well known to work (although it is more common to use Q1 for both START and STOP and to limit Q2 to ramp discharge duties
Re: [time-nuts] Regulating a pendulum clock
Advisable given the required mass will probably be in the 10-100 ton range. Bruce J. L. Trantham, M. D. wrote: Personally, I would get out of the way. : ) Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com]on Behalf Of Ian Sheffield Sent: Monday, August 09, 2010 1:17 PM To: j...@quik.com; 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] Regulating a pendulum clock What happens when the rope breaks? -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of J. Forster Sent: 09 August 2010 19:10 To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Regulating a pendulum clock You could put a large mass of concrete or somehing above the clock and crank it up and down, to balance out the computed gravity changes. :) -John == Unfortunately Gravity is not constant. Pendulum clocks show cyclic errors due to the influences of the Moon's and Sun's Gravitational fields. I forget the amounts but it is in the region of parts in 10 to the 7, which is easily measurable. This limits the compensations one can put into a pendulum clock. -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of mike cook Sent: 09 August 2010 18:21 To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Regulating a pendulum clock Le 09/08/2010 18:46, Bob Holmstrom a écrit : Food for thought. I find it interesting that no one has suggested alternatives to improving the performance of a pendulum clock other than controlling it with a higher performance clock. If the goal is a better clock why not attempt to understand the source of the errors and work on methods to control or compensate for them? Teddy Hall has been taken to task for using a quartz controlled oscillator to measure the amplitude of a pendulum in the control loop of his Littlemore clock. Tom Van Baak has developed techniques for analyzing the performance and hence potential error sources of pendulum clocks - perhaps he will share some of his work here. Horological history is full of many attempts at solutions to the problem, but it would seem that the creativity of this group might generate some new ideas that are more in the spirit of better timekeeping than attaching the pendulum to a better oscillator. How about a wireless controlled device attached to the pendulum that changes its position based on error sensor readings, not time errors, but instead, temperature, barometric pressure, gravity, etc. that would maintain a more constant pendulum period? Yup. We have temperature and pressure ICs available , I think that gravity is pretty constant if the clock isn't being moved about. Humididty might also need logging aswell. So it should be easy enough to predict the pendulums response to changes given a reasonable time of observation. That said, clocks have always been adjusted against better references.. IIRC Harrison (and probably others) was using star transits to regulate his long case clocks. Bob Holmström Editor Horological Science Newsletter www.hsn161.com ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
The attached single ended inverting driver is perhaps a better choice as it allows a dc coupled noniverting amplifier with gain and significant offset and drift to be substituted for the LT1010 buffer depicted if the frequency compensation is adjusted to suit. The series RC across the coil damps the coil resonance and the 1nF caps approximate wiring and coil capacitance to ground. A 10nF coil shunt capacitance and a series R of 400 ohms is included in the model. In practice the compensation should be adjusted to suit the actual coil used. A dc coupled discrete (or IC) audio power amplifier is one option for the noninverting amplifier. The noninverting amplifier may also have higher supply rails should this be useful/necessary. Bruce Bruce Griffiths wrote: A high voltage opamp (or a low voltage opamp with a discrete output stage with a voltage gain of at least 2) with -3V and + 30V supplies is perhaps the simplest method. The opamp merely senses the current flowing in a current sensing resistor and regulates this voltage drop to equal the output of a DAC. Alternatively it should be feasible to use a pair of opamps (plus output buffers) configured in a bridge arrangement to drive the coil from a single 30V supply. If one end of the coil has to remain near ground then a unity gain difference amplifier (with a discrete buffer with voltage gain) could be employed to implement a current source. A difference amplifier could also be employed together with an opamp (plus unity voltage gain discrete ouput stage) inverter to drive the coil from a single 30V supply. Bruce J. Forster wrote: Since it's inside a closed loop, the design is uncritical. One option is a high voltage Op-Amp with +/- 25 to 30 VDC supplies. You would set the OA gain to about 10, so 2.5 V in would yield 25 V out. and sum in a negative offset voltage so that +2.5 from the DAC yields 0.0 V out. I'd use something like a 100 K FB resistor and a 10K from the DAC, assuming it's a voltage output DAC. A 1 M to the -25 V supply would provide the 2.5 V offset. Another option would be to use two series opamps with the first set up as above, and the second as a unity gain inverter with input connected to the output of the first. The coil would connect between the two OA outputs. As one output swings high, the other mirrors that and goes low (just as in an H bridge). Stability might be an issue, but this has the advantage of only needing a +/- 15 supplies. FWIW, -John = Hi all, I have a Seimens master clock with a Reiffler pendulum. A lovely piece of work that used to provide time services in the 40s. Being a master clock it has contacts that open and close on each pendulum swing and so I can monitor it's accuracy quite easily using gps and my 5370B. I've adjusted it as best I can and the best I can get is about 50 ms over 24 hours. However that was a one off. Temp and air pressure cause variations of up to 300 ms and it changes direction too. Basically it's hard to keep accurate. It also has a coil mounted near the pendulum and a fixed magnet on the pendulum bar and this coil connects to a box down below with a meter and a knob. They are labelled in sec/day. The electronics in the box are not clear (being quite old) but by measuring the current in the coil it quite simply increases the current one way to slow the clock and the other way to speed it up. (I'll admit the physics of this doesn't make sense to me - but it works!) It's about 25v in the coil and goes up to 60mA max. Even at levels of 2mA has an effect. Using this control it's quite easy to manually bring the clock back to the right time if it's say half a second fast. What I want to do is control the current in the coil with a micro controller which I have attached to a rubidium oscillator. Getting the pps from the pendulum clock in and comparing to actual time is easy, but I need a way to control the current through the coil so it can dynamically adjust the clock. I need the current to go from say -10 to +10 mA (at 25v) and this needs to be controlled via a micro controller output (which goes from 0 to 5 with 2.5 being the 0mA point). I can either use the D/A in the controller (or PWM an output I suppose). I'd appreciate some thoughts on circuits to do this. Software side is not a problem. Jim Palfreyman ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions
Re: [time-nuts] Regulating a pendulum clock
J. Forster wrote: You are picking very unimportant nits. If there were a small noise spike from the opamp, it'd goose the pendulum a tiny amount. That would be corrected on the next swing. Heuristic analysis of this type is counter productive. You are turning a trip to the corner store into an Apollo Moon Mission. Reliability is paramount in a circuit that may be required to work for decades. BTW, since the =drive does not to be bipolar, one of the NPN and PNP transistors can be deleted. They never turn on. So you are left with two opamsa, =each with a simple emitter follower. The original request was for a bipolar drive. The lack of short circuit protection is poor design practice when driving an external load. -John == Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
No protection against external shorts or other undesired events. Extensive analog filtering to avoid creating an effective radiator of noise may also be necessary. Simple analog techniques are probably simpler/cheaper once the necessary filtering and protection are included. Bruce Don Latham wrote: Hmmm lemme see. I think I'd use a 12 volt supply and two transistors driven by two outputs on my Arduino,basic stamp,picaxe or other whizzie. I'd then implement a PID controller essentially using the 1 sec pulse from the pendulum and the 1 sec pulse from my Rb, satellite receiver, crystal clock, or whatever. The appropriate output pin will be brought to ground, and the other driven as a pdf with the rate given by the pid loop. Temperature and even pressure corrections can be applied within the gizzie software. External parts, minimum. Opportunity to play with tuning, maximum. Don Bruce Griffiths J. Forster wrote: You are picking very unimportant nits. If there were a small noise spike from the opamp, it'd goose the pendulum a tiny amount. That would be corrected on the next swing. Heuristic analysis of this type is counter productive. You are turning a trip to the corner store into an Apollo Moon Mission. Reliability is paramount in a circuit that may be required to work for decades. BTW, since the =drive does not to be bipolar, one of the NPN and PNP transistors can be deleted. They never turn on. So you are left with two opamsa, =each with a simple emitter follower. The original request was for a bipolar drive. The lack of short circuit protection is poor design practice when driving an external load. -John == Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
Mere fast blow fuses aren't usually precise enough to protect transistors against over current unless one uses rather large transistors. Overcurrent protected drivers are available and readily designed/built. Protection against di/dt transients due to external events is also advisable. Minimising the parts count isn't necessarily conducive to improved reliability when external hazards aren't taken into account. Merely resonating the coil without other filtering doesnt necessarily lead to low EMI when driving it with a voltage waveform having high edge slew rates. Some edge filtering to control the current flowing in the load capacitance is also advisable. Bruce Don Latham wrote: fast blow fuse, resonate the coil to the pwm frequency. Parts count small, tinkering in software instead of breathing lead fumes or whatever noxious stuff the Europeans have forced us to use... Don Bruce Griffiths No protection against external shorts or other undesired events. Extensive analog filtering to avoid creating an effective radiator of noise may also be necessary. Simple analog techniques are probably simpler/cheaper once the necessary filtering and protection are included. Bruce Don Latham wrote: Hmmm lemme see. I think I'd use a 12 volt supply and two transistors driven by two outputs on my Arduino,basic stamp,picaxe or other whizzie. I'd then implement a PID controller essentially using the 1 sec pulse from the pendulum and the 1 sec pulse from my Rb, satellite receiver, crystal clock, or whatever. The appropriate output pin will be brought to ground, and the other driven as a pdf with the rate given by the pid loop. Temperature and even pressure corrections can be applied within the gizzie software. External parts, minimum. Opportunity to play with tuning, maximum. Don Bruce Griffiths J. Forster wrote: You are picking very unimportant nits. If there were a small noise spike from the opamp, it'd goose the pendulum a tiny amount. That would be corrected on the next swing. Heuristic analysis of this type is counter productive. You are turning a trip to the corner store into an Apollo Moon Mission. Reliability is paramount in a circuit that may be required to work for decades. BTW, since the =drive does not to be bipolar, one of the NPN and PNP transistors can be deleted. They never turn on. So you are left with two opamsa, =each with a simple emitter follower. The original request was for a bipolar drive. The lack of short circuit protection is poor design practice when driving an external load. -John == Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
A high voltage opamp (or a low voltage opamp with a discrete output stage with a voltage gain of at least 2) with -3V and + 30V supplies is perhaps the simplest method. The opamp merely senses the current flowing in a current sensing resistor and regulates this voltage drop to equal the output of a DAC. Alternatively it should be feasible to use a pair of opamps (plus output buffers) configured in a bridge arrangement to drive the coil from a single 30V supply. If one end of the coil has to remain near ground then a unity gain difference amplifier (with a discrete buffer with voltage gain) could be employed to implement a current source. A difference amplifier could also be employed together with an opamp (plus unity voltage gain discrete ouput stage) inverter to drive the coil from a single 30V supply. Bruce Jim Palfreyman wrote: Hi all, I have a Seimens master clock with a Reiffler pendulum. A lovely piece of work that used to provide time services in the 40s. Being a master clock it has contacts that open and close on each pendulum swing and so I can monitor it's accuracy quite easily using gps and my 5370B. I've adjusted it as best I can and the best I can get is about 50 ms over 24 hours. However that was a one off. Temp and air pressure cause variations of up to 300 ms and it changes direction too. Basically it's hard to keep accurate. It also has a coil mounted near the pendulum and a fixed magnet on the pendulum bar and this coil connects to a box down below with a meter and a knob. They are labelled in sec/day. The electronics in the box are not clear (being quite old) but by measuring the current in the coil it quite simply increases the current one way to slow the clock and the other way to speed it up. (I'll admit the physics of this doesn't make sense to me - but it works!) It's about 25v in the coil and goes up to 60mA max. Even at levels of 2mA has an effect. Using this control it's quite easy to manually bring the clock back to the right time if it's say half a second fast. What I want to do is control the current in the coil with a micro controller which I have attached to a rubidium oscillator. Getting the pps from the pendulum clock in and comparing to actual time is easy, but I need a way to control the current through the coil so it can dynamically adjust the clock. I need the current to go from say -10 to +10 mA (at 25v) and this needs to be controlled via a micro controller output (which goes from 0 to 5 with 2.5 being the 0mA point). I can either use the D/A in the controller (or PWM an output I suppose). I'd appreciate some thoughts on circuits to do this. Software side is not a problem. Jim Palfreyman ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
The 60mA load current would be problematic for most common opamps without an output buffer stage. High voltage opamps are relatively rare. Bruce J. Forster wrote: Since it's inside a closed loop, the design is uncritical. One option is a high voltage Op-Amp with +/- 25 to 30 VDC supplies. You would set the OA gain to about 10, so 2.5 V in would yield 25 V out. and sum in a negative offset voltage so that +2.5 from the DAC yields 0.0 V out. I'd use something like a 100 K FB resistor and a 10K from the DAC, assuming it's a voltage output DAC. A 1 M to the -25 V supply would provide the 2.5 V offset. Another option would be to use two series opamps with the first set up as above, and the second as a unity gain inverter with input connected to the output of the first. The coil would connect between the two OA outputs. As one output swings high, the other mirrors that and goes low (just as in an H bridge). Stability might be an issue, but this has the advantage of only needing a +/- 15 supplies. FWIW, -John = Hi all, I have a Seimens master clock with a Reiffler pendulum. A lovely piece of work that used to provide time services in the 40s. Being a master clock it has contacts that open and close on each pendulum swing and so I can monitor it's accuracy quite easily using gps and my 5370B. I've adjusted it as best I can and the best I can get is about 50 ms over 24 hours. However that was a one off. Temp and air pressure cause variations of up to 300 ms and it changes direction too. Basically it's hard to keep accurate. It also has a coil mounted near the pendulum and a fixed magnet on the pendulum bar and this coil connects to a box down below with a meter and a knob. They are labelled in sec/day. The electronics in the box are not clear (being quite old) but by measuring the current in the coil it quite simply increases the current one way to slow the clock and the other way to speed it up. (I'll admit the physics of this doesn't make sense to me - but it works!) It's about 25v in the coil and goes up to 60mA max. Even at levels of 2mA has an effect. Using this control it's quite easy to manually bring the clock back to the right time if it's say half a second fast. What I want to do is control the current in the coil with a micro controller which I have attached to a rubidium oscillator. Getting the pps from the pendulum clock in and comparing to actual time is easy, but I need a way to control the current through the coil so it can dynamically adjust the clock. I need the current to go from say -10 to +10 mA (at 25v) and this needs to be controlled via a micro controller output (which goes from 0 to 5 with 2.5 being the 0mA point). I can either use the D/A in the controller (or PWM an output I suppose). I'd appreciate some thoughts on circuits to do this. Software side is not a problem. Jim Palfreyman ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
A high voltage opamp (or a low voltage opamp with a discrete output stage with a voltage gain of at least 2) with -3V and + 30V supplies is perhaps the simplest method. The opamp merely senses the current flowing in a current sensing resistor and regulates this voltage drop to equal the output of a DAC. Alternatively it should be feasible to use a pair of opamps (plus output buffers) configured in a bridge arrangement to drive the coil from a single 30V supply. If one end of the coil has to remain near ground then a unity gain difference amplifier (with a discrete buffer with voltage gain) could be employed to implement a current source. A difference amplifier could also be employed together with an opamp (plus unity voltage gain discrete ouput stage) inverter to drive the coil from a single 30V supply. Bruce J. Forster wrote: Since it's inside a closed loop, the design is uncritical. One option is a high voltage Op-Amp with +/- 25 to 30 VDC supplies. You would set the OA gain to about 10, so 2.5 V in would yield 25 V out. and sum in a negative offset voltage so that +2.5 from the DAC yields 0.0 V out. I'd use something like a 100 K FB resistor and a 10K from the DAC, assuming it's a voltage output DAC. A 1 M to the -25 V supply would provide the 2.5 V offset. Another option would be to use two series opamps with the first set up as above, and the second as a unity gain inverter with input connected to the output of the first. The coil would connect between the two OA outputs. As one output swings high, the other mirrors that and goes low (just as in an H bridge). Stability might be an issue, but this has the advantage of only needing a +/- 15 supplies. FWIW, -John = Hi all, I have a Seimens master clock with a Reiffler pendulum. A lovely piece of work that used to provide time services in the 40s. Being a master clock it has contacts that open and close on each pendulum swing and so I can monitor it's accuracy quite easily using gps and my 5370B. I've adjusted it as best I can and the best I can get is about 50 ms over 24 hours. However that was a one off. Temp and air pressure cause variations of up to 300 ms and it changes direction too. Basically it's hard to keep accurate. It also has a coil mounted near the pendulum and a fixed magnet on the pendulum bar and this coil connects to a box down below with a meter and a knob. They are labelled in sec/day. The electronics in the box are not clear (being quite old) but by measuring the current in the coil it quite simply increases the current one way to slow the clock and the other way to speed it up. (I'll admit the physics of this doesn't make sense to me - but it works!) It's about 25v in the coil and goes up to 60mA max. Even at levels of 2mA has an effect. Using this control it's quite easy to manually bring the clock back to the right time if it's say half a second fast. What I want to do is control the current in the coil with a micro controller which I have attached to a rubidium oscillator. Getting the pps from the pendulum clock in and comparing to actual time is easy, but I need a way to control the current through the coil so it can dynamically adjust the clock. I need the current to go from say -10 to +10 mA (at 25v) and this needs to be controlled via a micro controller output (which goes from 0 to 5 with 2.5 being the 0mA point). I can either use the D/A in the controller (or PWM an output I suppose). I'd appreciate some thoughts on circuits to do this. Software side is not a problem. Jim Palfreyman ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. attachment: Coil_Driver.gif___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulating a pendulum clock
J. Forster wrote: OK. You know better. BTW, op-amp noise is essentially irrelevant in this application, and the C's across the FB resistors limit slew rates so there is no significant dI/dt to cause voltage spikes. Noise is never irrelevant. You havent shown that its insignificant either. In the real world such dv/dt assumptions with inductive loads lead to fried parts. For example if the circuit oscillates at high frequency because the compensation isnt correct/effective or the feedback wire becomes detached or the power supply goes down suddently due to a crowbar event then high dv/dt at the opamp/buffer output is possible. -John Bruce Your naive stabilisation scheme wont work, try simulating it. 741's are somewhat noisier than necessary. Omitting the diodes with an inductive load almost inevitably leads to transistor or opamp destruction. Bruce J. Forster wrote: IMO, far too complicated. I'd use a series pair of u741s each with a complementary emitter follower. 2 u741s, 2x 2N2102, 2x 2N4036, 5 resistors. Maybe 2x .01 caos to stabilize the thing - |\| |---|c DAC --o--| \ | |\ 2N2102 | | / --o-o |--C R |/| | |/ 2N4036 || | |---|c || || |o-to input of mirror image Best, -J = The attached circuit schematic illustrates the Howland current source plus inverting amplifier drive technique. It also illustrates a method of frequency compensation (series RC connected across the coil). Of course one can either use discrete buffers or high current opamps. However for improved accuracy using a difference amplifier with built in pretrimmed resistors for the Howland current source may be preferable, in which case a discrete buffer stage or equivalent may be required. Bruce J. Forster wrote: There are cheap, split supply audio amp ICs that'd work, or you could use a u741 with a complementary-symmetry output buffer of discrete transistors. Crossover distortion would be essentially irrelevant, keeping the parts count very low. -John The 60mA load current would be problematic for most common opamps without an output buffer stage. High voltage opamps are relatively rare. Bruce J. Forster wrote: Since it's inside a closed loop, the design is uncritical. One option is a high voltage Op-Amp with +/- 25 to 30 VDC supplies. You would set the OA gain to about 10, so 2.5 V in would yield 25 V out. and sum in a negative offset voltage so that +2.5 from the DAC yields 0.0 V out. I'd use something like a 100 K FB resistor and a 10K from the DAC, assuming it's a voltage output DAC. A 1 M to the -25 V supply would provide the 2.5 V offset. Another option would be to use two series opamps with the first set up as above, and the second as a unity gain inverter with input connected to the output of the first. The coil would connect between the two OA outputs. As one output swings high, the other mirrors that and goes low (just as in an H bridge). Stability might be an issue, but this has the advantage of only needing a +/- 15 supplies. FWIW, -John = Hi all, I have a Seimens master clock with a Reiffler pendulum. A lovely piece of work that used to provide time services in the 40s. Being a master clock it has contacts that open and close on each pendulum swing and so I can monitor it's accuracy quite easily using gps and my 5370B. I've adjusted it as best I can and the best I can get is about 50 ms over 24 hours. However that was a one off. Temp and air pressure cause variations of up to 300 ms and it changes direction too. Basically it's hard to keep accurate. It also has a coil mounted near the pendulum and a fixed magnet on the pendulum bar and this coil connects to a box down below with a meter and a knob. They are labelled in sec/day. The electronics in the box are not clear (being quite old) but by measuring the current in the coil it quite simply increases the current one way to slow the clock and the other way to speed it up. (I'll admit the physics of this doesn't make sense to me - but it works!) It's about 25v in the coil and goes up to 60mA max. Even at levels of 2mA has an effect. Using this control it's quite easy to manually bring the clock back to
Re: [time-nuts] Buffer / distribution amplifier for TCXO
Henry Hallam wrote: Dear time nuts, Background: I have built a GPS receiver based around the SE4120L front end IC [1]. I used a KT3225 TCXO [2] at 16.3676MHz driving the front end through a 10nF series capacitor as in the example circuit in [1]. Inside the front end, this oscillator is multiplied up to form a local oscillator at 1571.2896 MHz. The 16.3676MHz signal is also divided to form a 4.0919MHz sampling clock. Digital I and Q samples then go to a DSP where the GPS signal processing is done in software. My receiver works nicely, getting it online was a boatload of fun and I'm hoping to make it available soon along with open-source software as a GPS experimenter's kit. Problem: I'd like to clock multiple receivers from a single 16.3676MHz oscillator, in order to combine measurements from multiple antennas. The clocks must be at the same frequency, i.e. from the same source, but it is not necessary that they have any particular phase relationship as phase offsets are removed in the navigation processing. What sort of distribution amplifier should I use to split the output of one TCXO into four front ends? Do I need some kind of impedance matching network? How would I go about designing that? This sort of analog/RF design is unfamiliar territory for me, though I'd like to learn. The TCXO advertises a minimum output level of 0.8Vpp into (10kohm in parallel with 10pF). The front end requires a minimum oscillator drive level of 0.2Vpp. The front end datasheet lists recommended crystal parameters including a load capacitance of 10pF (typ), although I don't know whether or not that refers to the front end input capacitance. My guess is that phase noise performance is not particularly crucial, at least by time-nuts standards. I guess it would be nice if the amplifier didn't make the phase noise significantly worse than it already is from the cheap TCXO. Many thanks, Henry Hallam [1] http://www.sige.com/support/download-form.html?dl=DST-00059_SE4120L_Datasheet_Rev_3p5_CYW_May-26-2009.pdf [2] http://global.kyocera.com/prdct/electro/pdf/tcxo/172_e.pdf ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. The TCXO output waveform is presumably a clipped sinewave as required by the SE4120L? In which case a linear distribution amplifier is probably required. With only a ~3V supply available, options for the distribution amplifier topology are somewhat limited. In principle you could use an emitter follower driving 4 other emitter followers with a resistor in series with the emitters of the output devices and the AC coupled loads to match the source to the interconnecting cable impedance to minimise reflections without requiring excessive dissipation in the emitter followers. With the low voltage supply available, using an RF choke is series with the emitter follower's emitter to ground resistor will also be useful in achieving the required dynamic range. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Buffer / distribution amplifier for TCXO
The GPS receiver chip actually specifies that a clipped sinewave should be used. Presumably this is necessary to limit the harmonic contents. In which case low pass filtering the CMOS outputs may be necessary. The 74AHC04 or equivalent may be a better choice as its ground and Vcc bounce is lower than that of a 74AC04. Bruce Bob Camp wrote: Hi I suspect you will find that the phase noise floor of the distribution system does indeed matter. Likely the easy way to go: Square the TCXO up with a biased CMOS inverter (at least as fast as a 74AC04). Run a seperate inverter to drive each of the receivers. A hex inverter chip would do it all quite nicely. There should be plenty of isolation and far more signal than is needed. Attenuating it at the receiver with a pair of resistors should get all the levels to match up. If you want to get fancy, transformer couple into each receiver after attenuating. Bob -- From: Henry Hallam he...@pericynthion.org Sent: Wednesday, August 04, 2010 1:46 PM To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: [time-nuts] Buffer / distribution amplifier for TCXO Dear time nuts, Background: I have built a GPS receiver based around the SE4120L front end IC [1]. I used a KT3225 TCXO [2] at 16.3676MHz driving the front end through a 10nF series capacitor as in the example circuit in [1]. Inside the front end, this oscillator is multiplied up to form a local oscillator at 1571.2896 MHz. The 16.3676MHz signal is also divided to form a 4.0919MHz sampling clock. Digital I and Q samples then go to a DSP where the GPS signal processing is done in software. My receiver works nicely, getting it online was a boatload of fun and I'm hoping to make it available soon along with open-source software as a GPS experimenter's kit. Problem: I'd like to clock multiple receivers from a single 16.3676MHz oscillator, in order to combine measurements from multiple antennas. The clocks must be at the same frequency, i.e. from the same source, but it is not necessary that they have any particular phase relationship as phase offsets are removed in the navigation processing. What sort of distribution amplifier should I use to split the output of one TCXO into four front ends? Do I need some kind of impedance matching network? How would I go about designing that? This sort of analog/RF design is unfamiliar territory for me, though I'd like to learn. The TCXO advertises a minimum output level of 0.8Vpp into (10kohm in parallel with 10pF). The front end requires a minimum oscillator drive level of 0.2Vpp. The front end datasheet lists recommended crystal parameters including a load capacitance of 10pF (typ), although I don't know whether or not that refers to the front end input capacitance. My guess is that phase noise performance is not particularly crucial, at least by time-nuts standards. I guess it would be nice if the amplifier didn't make the phase noise significantly worse than it already is from the cheap TCXO. Many thanks, Henry Hallam [1] http://www.sige.com/support/download-form.html?dl=DST-00059_SE4120L_Datasheet_Rev_3p5_CYW_May-26-2009.pdf [2] http://global.kyocera.com/prdct/electro/pdf/tcxo/172_e.pdf ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Buffer / distribution amplifier for TCXO
Henry Hallam wrote: On Wed, Aug 4, 2010 at 1:50 PM, Bruce Griffiths bruce.griffi...@xtra.co.nz wrote: The TCXO output waveform is presumably a clipped sinewave as required by the SE4120L? I posted the waveform at http://www.pericynthion.org/stuff/KT3225_500mV_per_div.jpg Does that count as clipped sine? If not, it seems to work anyway. Its something like a clipped sine albeit with some ringing as it is clipped. It more closely resembles a low pass filtered square wave. In which case a linear distribution amplifier is probably required. With only a ~3V supply available, options for the distribution amplifier topology are somewhat limited. I'm making a custom board that will include the TCXO and distribution amplifier (as well as some digital stuff to allow the 4 receivers to communicate), so it can have whatever power supplies it needs. In principle you could use an emitter follower driving 4 other emitter followers with a resistor in series with the emitters of the output devices and the AC coupled loads to match the source to the interconnecting cable impedance to minimise reflections without requiring excessive dissipation in the emitter followers. With the low voltage supply available, using an RF choke is series with the emitter follower's emitter to ground resistor will also be useful in achieving the required dynamic range. Thanks. Henry Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Buffer / distribution amplifier for TCXO
Is that also true for AHC devices which otherwise have similar characteristics (apart from ground bounce) to AC devices? Bruce Bob Camp wrote: Hi The phase noise floor of the HC is *much* higher than the floor of the AC gates. The main reason it specifies clipped sine is that's what the cheap TCXO's put out. Bob On Aug 4, 2010, at 6:42 PM, Bruce Griffiths wrote: The GPS receiver chip actually specifies that a clipped sinewave should be used. Presumably this is necessary to limit the harmonic contents. In which case low pass filtering the CMOS outputs may be necessary. The 74AHC04 or equivalent may be a better choice as its ground and Vcc bounce is lower than that of a 74AC04. Bruce Bob Camp wrote: Hi I suspect you will find that the phase noise floor of the distribution system does indeed matter. Likely the easy way to go: Square the TCXO up with a biased CMOS inverter (at least as fast as a 74AC04). Run a seperate inverter to drive each of the receivers. A hex inverter chip would do it all quite nicely. There should be plenty of isolation and far more signal than is needed. Attenuating it at the receiver with a pair of resistors should get all the levels to match up. If you want to get fancy, transformer couple into each receiver after attenuating. Bob -- From: Henry Hallamhe...@pericynthion.org Sent: Wednesday, August 04, 2010 1:46 PM To: Discussion of precise time and frequency measurementtime-nuts@febo.com Subject: [time-nuts] Buffer / distribution amplifier for TCXO Dear time nuts, Background: I have built a GPS receiver based around the SE4120L front end IC [1]. I used a KT3225 TCXO [2] at 16.3676MHz driving the front end through a 10nF series capacitor as in the example circuit in [1]. Inside the front end, this oscillator is multiplied up to form a local oscillator at 1571.2896 MHz. The 16.3676MHz signal is also divided to form a 4.0919MHz sampling clock. Digital I and Q samples then go to a DSP where the GPS signal processing is done in software. My receiver works nicely, getting it online was a boatload of fun and I'm hoping to make it available soon along with open-source software as a GPS experimenter's kit. Problem: I'd like to clock multiple receivers from a single 16.3676MHz oscillator, in order to combine measurements from multiple antennas. The clocks must be at the same frequency, i.e. from the same source, but it is not necessary that they have any particular phase relationship as phase offsets are removed in the navigation processing. What sort of distribution amplifier should I use to split the output of one TCXO into four front ends? Do I need some kind of impedance matching network? How would I go about designing that? This sort of analog/RF design is unfamiliar territory for me, though I'd like to learn. The TCXO advertises a minimum output level of 0.8Vpp into (10kohm in parallel with 10pF). The front end requires a minimum oscillator drive level of 0.2Vpp. The front end datasheet lists recommended crystal parameters including a load capacitance of 10pF (typ), although I don't know whether or not that refers to the front end input capacitance. My guess is that phase noise performance is not particularly crucial, at least by time-nuts standards. I guess it would be nice if the amplifier didn't make the phase noise significantly worse than it already is from the cheap TCXO. Many thanks, Henry Hallam [1] http://www.sige.com/support/download-form.html?dl=DST-00059_SE4120L_Datasheet_Rev_3p5_CYW_May-26-2009.pdf [2] http://global.kyocera.com/prdct/electro/pdf/tcxo/172_e.pdf ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Buffer / distribution amplifier for TCXO
Bruce Griffiths wrote: In which case a linear distribution amplifier is probably required. With only a ~3V supply available, options for the distribution amplifier topology are somewhat limited. In principle you could use an emitter follower driving 4 other emitter followers with a resistor in series with the emitters of the output devices and the AC coupled loads to match the source to the interconnecting cable impedance to minimise reflections without requiring excessive dissipation in the emitter followers. With the low voltage supply available, using an RF choke is series with the emitter follower's emitter to ground resistor will also be useful in achieving the required dynamic range. Bruce A more efficient buffer amplifier circuit schematic is attached. The series transformer feedback stage has high input impedance and an output impedance matched to the transmission line (yes it works well with long transmission lines as well). However a trifilar wound RF transformer is required. In principle the various GPS receivers could be connected to taps along an end terminated transmission line using feedthrough connections with compensation for the tap shunt capacitance if necessary. A lower impedance line (eg 50 ohms) could also be driven at the expense of a higher collector current. In this case the value of R3 would need to be reduced to around 100 ohms or so. Bruce attachment: TCXO_BUFFER.gif___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] buffer amp transformers...
For this application you'll need a bandwidth of somewhat more than 10MHz to preserve the clock slew rate. Those transformers are better suited to sinewave operation at 5 or 10MHz. If one uses a pair of transformers (one for the feedback and one to isolate the output) then wider bandwidth 1:1 transformers can be used. Or one could just elect to capacitively couple the load. Alternatively one can just wind one's own trifilar transformer using a suitable binocular ferrite core. Bruce Henry Hallam wrote: The spec sheet lists them as being good to 10MHz; would they be ok at 16MHz with a little more loss, or should I worry about resonances with parasitic capacitance? 73 de Henry M0HMH in Santa Cruz On Wed, Aug 4, 2010 at 9:14 PM,k6...@comcast.net wrote: I think I'm a time-nut; as symptoms I include (1) a lot of Mini-Circuits parts on my bench, (2) searches on eBay for Mini-Circuits goodies, and (3) the desire to know how my LPRO, 10811, and Thunderbolt are different, and how much better a Thunderbolt would be with a 10811 double-oven in it... Anyway, here's an eBay auction for 25 T-626 1:1:1 transformers -- item number: 220544907085 http://cgi.ebay.com/25-Mini-Circuits-T-626-RF-Transformers-0-01-10-MHz-/220544907085?cmd=ViewItempt=LH_DefaultDomain_0hash=item335980374d which look like just the thing for this amp... 73 de bob k6rtm in silicon valley - Message: 4 Date: Thu, 05 Aug 2010 10:05:39 +1200 From: Bruce Griffithsbruce.griffi...@xtra.co.nz Subject: Re: [time-nuts] Buffer / distribution amplifier for TCXO To: Discussion of precise time and frequency measurement time-nuts@febo.com Message-ID:4c59e433.6000...@xtra.co.nz Content-Type: text/plain; charset=iso-8859-1; Format=flowed Bruce Griffiths wrote: In which case a linear distribution amplifier is probably required. With only a ~3V supply available, options for the distribution amplifier topology are somewhat limited. In principle you could use an emitter follower driving 4 other emitter followers with a resistor in series with the emitters of the output devices and the AC coupled loads to match the source to the interconnecting cable impedance to minimise reflections without requiring excessive dissipation in the emitter followers. With the low voltage supply available, using an RF choke is series with the emitter follower's emitter to ground resistor will also be useful in achieving the required dynamic range. Bruce A more efficient buffer amplifier circuit schematic is attached. The series transformer feedback stage has high input impedance and an output impedance matched to the transmission line (yes it works well with long transmission lines as well). However a trifilar wound RF transformer is required. In principle the various GPS receivers could be connected to taps along an end terminated transmission line using feedthrough connections with compensation for the tap shunt capacitance if necessary. A lower impedance line (eg 50 ohms) could also be driven at the expense of a higher collector current. In this case the value of R3 would need to be reduced to around 100 ohms or so. Bruce -- next part -- A non-text attachment was scrubbed... Name: TCXO_BUFFER.gif Type: image/gif Size: 7990 bytes Desc: not available URL:http://www.febo.com/pipermail/time-nuts/attachments/20100805/091e6a74/attachment.gif -- ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts End of time-nuts Digest, Vol 73, Issue 12 * ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Updated Shera controller
You can't predict the settling time of an opamp from its slew rate or its gain-bandwidth product. The TS272 datasheet has no settling time spec whatsoever. In this case, since there is no spec it needs to be measured. Opamps with 2 or more cascaded gain stages like these are notorious for poor settling times. 10us is merely guesswork. The settling time could well be much longer and it may depend on the input signal level. Bruce Richard H McCorkle wrote: FYI, The TS272/TS274 have a slew rate of 5.5v/us at unity gain, the max voltage on the cap is 2.7v in the new design, and the voltage is read 10us after sample complete, so the buffer should have time to stabilize after the sample before being read. Richard Bruce wrote: Not really its both overkill as it doesnt timestamp, it only measures a time interval and underkill in that theres no DAC. There are also some concerns about the settling time of the TAC buffer opamp which isnt strictly necessary for the lower resolution required in this application. PPS timestamping only needs a single TAC. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Updated Shera controller
Instead of copying the Shera controller albeit with higher resolution its probably more cost effective to choose a microprocessor with built in time stamping capability. The PIC24FJ128GA for example allows 30ns timestamping resolution via its external timer capture inputs. Other microprocessors (or DSPs) are available with even higher timestamping resolution. Since modern GPS timing receivers can have a timing noise of a few ns (after sawtooth correction) it may also be useful to use a simple time to digital converter (eg TAC + ADC) to achieve a resolution of 1ns or so. Several microprocessors have built in ADCs with sufficient resolution and low leakage to allow trimpots, external buffer opamps etc to be dispensed with. The trimpots being replaced by interleaved software calibration. The microprocessor initiates a TAC calibration cycle after each external PPS event is timestamped. The resultant sequence of calibration coefficients can then be filtered and used to correct the PPS fine time stamp sequence. If the microprocessor also has a couple of PWM outputs then these can be used to implement a high resolution synchronously filtered DAC The synchronously filtered DAC requires a stable reference, a couple of opamps, a few analog switches plus a few resistors and capacitors. Neither the resistors or capacitors need to be extremely close tolerance parts. Bruce ewkeh...@aol.com wrote: What could you help with? Bert In a message dated 7/28/2010 3:05:26 P.M. Eastern Daylight Time, wpxs...@gmail.com writes: Bert, I for one, would be interested in that. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Updated Shera controller
ewkeh...@aol.com wrote: One more comment $ 40 would cover every thing including PC board, May go up to 50 depending what D/A you use. Bert In a message dated 7/28/2010 4:47:46 P.M. Eastern Daylight Time, p...@phk.freebsd.dk writes: In message4c5092fa.2030...@xtra.co.nz, Bruce Griffiths writes: Instead of copying the Shera controller albeit with higher resolution its probably more cost effective to choose a microprocessor with built in time stamping capability. Uhm, isn't this exactly where you want to use the still-smelling-like-brand-new-car PICTIC II with a good DAC and a microcontroller ? Not really its both overkill as it doesnt timestamp, it only measures a time interval and underkill in that theres no DAC. There are also some concerns about the settling time of the TAC buffer opamp which isnt strictly necessary for the lower resolution required in this application. PPS timestamping only needs a single TAC. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Basic question regarding comparing two frequencies
Hal Murray wrote: There is another way to compare two frequencies, relevant when they are very close together. I divide a reference down to 100KHz and use it to clock a phase detector made of a pair of D flip flops. The unknown (divided to 100KHz) is fed into the circuit and an output that is proportional to the phase difference appears on the output as a changing mark-space ratio. I like it. Thanks. How did you pick 100 KHz? Using CMOS and a precise power supply (because under no load, CMOS output is precisely rail to rail), the averaged output (100ms RC filter) is fed to a strip chart recorder. Has anybody checked the edge cases and/or linearity of a setup like this? The recorder shows the changing phase difference and folds back each time a whole cycle passes. A 12 bit analog data logger resolves 2.5ns of phase and gives data for further analysis. Is 2.5 ns good enough? What would you gain by using a 16 bit DAC? A ratiometric ADC where the ADC uses the (low pass filtered) CMOS supply as its reference is probably advisable when using high resolution ADCs. A high resolution sigma delta ADC that aloows an external reference to be used may be useful for this application. If 2.5 ns is good enough, I'll bet you can do the whole thing in digital logic. Just get a fast FPGA/CPLD. I haven't done a serious design, but a quick check at some old data sheets shows it's not silly. You could probably bump it up by another factor of 2 with some external (p)ECL chips. If one used an FPGA with an internal 500MHz (use the internal PLL available in some FPGAs) clock and dual edge clocking or a 1GHz internal clock, 1ns resolution should be readily achievable. However it may be advisable to use something like LVDS inputs to alleviate the effects of ground and Vcc bounce. If you need more resolution then one could always sample the outputs of an internal tapped delay line using internal gates as delay elements. With a suitable FPGA a resolution of a few hundred ps is feasible. If the delay line delay is more than 1 clock period then an embedded calibration of the delay line is possible from the coarse (1ns) count and the fine count from the internal tapped delay line. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Info on MTI osc.
Luis Cupido wrote: MTI 230-0546-A Is the pinout similar to the other 230 series OCXOs?: http://www.mti-milliren.com/pdfs/230.pdf Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
Chris All the details are in the article: http://www.edn.com/contents/images/6607197.pdf However it would be best to read the article posted by Bob Camp first: Bruce Chris Stake wrote: Hi Bruce, This sounds like a promising idea, please could you expand on the synchronous filter technique? I have seen some articles about how such filters can be used to clean up the data from rotating machinery for vibration analysis etc. but I don't follow how they can be used in a PWM application. Regards Chris Stake Its possible to build a 24 bit resolution D/A using a synchronously filtered PWM circuit. A pair of PWM outputs and a few relatively low precision resistors and capacitors together with a low noise low drift reference are required. The technique takes advantage of the fact that the required EFC voltage changes slowly and isnt updated at a highg rate. The synchronous filter technique eliminates the very long time constant RC filters required with an asynchronously filtered PWM waveform. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
Here's a link to a pdf version of the synchronously filtered low ripple pwm dac: http://www.edn.com/contents/images/6553625.pdf Bruce Bruce Griffiths wrote: Chris All the details are in the article: http://www.edn.com/contents/images/6607197.pdf However it would be best to read the article posted by Bob Camp first: Bruce Chris Stake wrote: Hi Bruce, This sounds like a promising idea, please could you expand on the synchronous filter technique? I have seen some articles about how such filters can be used to clean up the data from rotating machinery for vibration analysis etc. but I don't follow how they can be used in a PWM application. Regards Chris Stake Its possible to build a 24 bit resolution D/A using a synchronously filtered PWM circuit. A pair of PWM outputs and a few relatively low precision resistors and capacitors together with a low noise low drift reference are required. The technique takes advantage of the fact that the required EFC voltage changes slowly and isnt updated at a highg rate. The synchronous filter technique eliminates the very long time constant RC filters required with an asynchronously filtered PWM waveform. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
Attila Kinali wrote: Moin, On Sat, 26 Jun 2010 21:14:02 EDT ewkeh...@aol.com wrote: What you want is basically a Shera Board. That design has been around for quite some time and has served me very well. Yes. The Shera Board and similar designs serve as an example for me. I have a total of six running including two controlling Rubidium. There are in my opinion a couple of problems: not every 4066 works on the design the 18 bit D/A is very hard to find and now expensive and the single step of the D/A is intended for a 1.7 E-13 frequency step. Yes. My goal is to update the venerable 4066 with something more modern and have components that are easy to get trough farnell, digikey, mouser, and all the other distributors. Yes, 16bit D/A seems to be the maximum that is currently available. It crossed my mind to build a 24bit R-2R D/A using discrete components, but this might have actually a worse performance than a off the shelf 16bit D/A. (temperature drifft, resistor values missmatch, EMI, etc) Attila Kinali Its possible to build a 24 bit resolution D/A using a synchronously filtered PWM circuit. A pair of PWM outputs and a few relatively low precision resistors and capacitors together with a low noise low drift reference are required. The technique takes advantage of the fact that the required EFC voltage changes slowly and isnt updated at a highg rate. The synchronous filter technique eliminates the very long time constant RC filters required with an asynchronously filtered PWM waveform. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
The problem is that the gain and offset of the 2 DACs changes with time and temperature so that the required corrections also change. Ideally an autocalibration technique would be used to dynamically track such changes. Since changes in the coarse DAC are only required infrequently and the mismatch only affects the region around coarse DAC transitions which are relatively infrequent (or should be) most designers choose to live with the increased loop settling time at such transitions. With sufficient overlap between the coarse and fine DACs only small fine DAC changes should be required to compensate for mismatch between the coarse and fine DACs after a change in the coarse DAC output. The coarse + fine DAC approach is used in some GPSDOs and in particle accelerator control systems. Bruce ewkeh...@aol.com wrote: Hi, just a clarification, I did write 4066 it is a 4046 that I replaced. Take a look at the MCP 4822 dual 12 bit D/A In the data sheet they have an example using one for coarse, the other for fine steps, I realize that the transition is not perfect but maybe code can compensate for the transition. Bert Kehren In a message dated 6/29/2010 5:10:39 A.M. Eastern Daylight Time, att...@kinali.ch writes: Moin, On Sat, 26 Jun 2010 21:14:02 EDT ewkeh...@aol.com wrote: What you want is basically a Shera Board. That design has been around for quite some time and has served me very well. Yes. The Shera Board and similar designs serve as an example for me. I have a total of six running including two controlling Rubidium. There are in my opinion a couple of problems: not every 4066 works on the design the 18 bit D/A is very hard to find and now expensive and the single step of the D/A is intended for a 1.7 E-13 frequency step. Yes. My goal is to update the venerable 4066 with something more modern and have components that are easy to get trough farnell, digikey, mouser, and all the other distributors. Yes, 16bit D/A seems to be the maximum that is currently available. It crossed my mind to build a 24bit R-2R D/A using discrete components, but this might have actually a worse performance than a off the shelf 16bit D/A. (temperature drifft, resistor values missmatch, EMI, etc) Attila Kinali -- If you want to walk fast, walk alone. If you want to walk far, walk together. -- African proverb ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
Attila Kinali wrote: On Tue, 29 Jun 2010 21:32:10 +1200 Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Its possible to build a 24 bit resolution D/A using a synchronously filtered PWM circuit. A pair of PWM outputs and a few relatively low precision resistors and capacitors together with a low noise low drift reference are required. The technique takes advantage of the fact that the required EFC voltage changes slowly and isnt updated at a highg rate. The synchronous filter technique eliminates the very long time constant RC filters required with an asynchronously filtered PWM waveform. I've thought about that, but i'm afraid that this will add too much phase noise trough EFC noise. Though, i have not calculated how much noise this would generate. Attila Kinali How do you conclude that? You don't know what the circuit is and you've never tested it. Ulrich has, and the output noise is very low. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
Bob Camp wrote: Hi Are you referring to something like this: http://www.electronicsweekly.com/Articles/2008/05/01/43680/fast-settling-syn chronous-pwm-dac-filter-has-almost-no.htm as a synchronous filter for the PWM? Bob Yes, that is the original article. There's a later one (the link is in the archives) which shows how to use a pair of 16 bit PWM signals in conjunction with such a filter. However there is an error in one of the resistor values. Ulrich built and tested a 24 bit version using a pair of 16 bit PWM signals. Bruce -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Bruce Griffiths Sent: Tuesday, June 29, 2010 4:07 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] yet another GPSDO design, or so Attila Kinali wrote: On Tue, 29 Jun 2010 21:32:10 +1200 Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Its possible to build a 24 bit resolution D/A using a synchronously filtered PWM circuit. A pair of PWM outputs and a few relatively low precision resistors and capacitors together with a low noise low drift reference are required. The technique takes advantage of the fact that the required EFC voltage changes slowly and isnt updated at a highg rate. The synchronous filter technique eliminates the very long time constant RC filters required with an asynchronously filtered PWM waveform. I've thought about that, but i'm afraid that this will add too much phase noise trough EFC noise. Though, i have not calculated how much noise this would generate. Attila Kinali How do you conclude that? You don't know what the circuit is and you've never tested it. Ulrich has, and the output noise is very low. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] yet another GPSDO design, or so
Hal Murray wrote: bruce.griffi...@xtra.co.nz said: Its possible to build a 24 bit resolution D/A using a synchronously filtered PWM circuit. A pair of PWM outputs and a few relatively low precision resistors and capacitors together with a low noise low drift reference are required. The technique takes advantage of the fact that the required EFC voltage changes slowly and isnt updated at a highg rate. The synchronous filter technique eliminates the very long time constant RC filters required with an asynchronously filtered PWM waveform. 24 bits is 16,777,216. At a reasonable clock rate, that's one second. Not if one uses a pair of 16bit PWM circuits to produce a DAC with 24 bit resolution. a few 0.1% resistors then suffice to achieve 24 bit linearity. Another approach is to distribute the individual bits rather than clump them together. If you want 1/2, send 10101010 rather than . You would have to do something like build a bit pattern in memory and use a serial port to send it out. With a synchronous filter the settling time (for small output changes) is equal to the PWM period. The synchronous filter uses a variation of a dual slope error integrator, the output of which when sampled is equal to the desired output. The effect of dielectric absorption in the error integrator can be reduced by implementing a mutislope integrator rather than a dual slope version. Its then possible to use a pair of 8 bit PWM signals to achieve 24 bit resolution. That shifts the frequency of the junk so that it's easier to filter out and/or reduces the amplitude. If you send 10101010, you have lots of energy but it's at 8 MHz. If you send 100, you have energy at 1 Hz, but it's only 1/1600 as big. Or something like that. [Since this is a linear system, you will get that spur with any odd number of 1s.] I can't determine if that's good enough. I think the math is similar to the spurs you get from a DDS. Simulated that, and Ulrich did some testing, the spurs can be problematic. The synchronous PWM circuit is much easier to filter as the synchronous output noise amplitude (with a constant input) due to sampling charge injection need not be more than a few microvolts. That is there is a small spur with an amplitude of a few microvolts at the PWM repetition rate. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Advantages Disadvantages of the TPLL Method
WarrenS wrote: Long explanations, cause I try to explain, the best I can, when I say something is WRONG or misleading Magnus Posted: EFC linearity will remain an issue for analog oscillators. The oscillator gain will differ depending on offset voltage and temperature. TRUE it is an issue, but somewhat misleading because it need NOT be a problem or limitation (mostly) EFC Linearity can be an issue because the TPLL is limited by the performance of the reference oscillator in lots of ways. BUT Oscillator EFC gain or linearity are not likely to be of much concern or a limitation for high end performance. The gain nonlinearity I've measured can vary two to one over the full range of a good Oscillator but it is more like 10% over the normally used range, if one stays well away from the end points. NOT so good but livable if you are not making something real accurate. BUT For all my accurate stuff, when using a HP 10811, I limit the full-scale change to 1e-9 or 1e-8 at most. This uses such a small part of the total EFC range, that the nonlinearity effects are generally below the noise level and of little concern at all. The fact that Oscillator gain does differ with the EFC voltage (offset voltage), means if you want to get max accuracy out of the TPLL, it will need to be calibrated at the EFC offset voltage it is being used at. One simple solution, if the OSC also has a independent manual Freq adjustment like the single oven 10811, is to use it always set the EFC voltage to be near zero volts. BTW calibration need not be much of a problem, because it can be a static calibration. If and only if injection locking isn't significant. This needs to be established for each setup. The simplest way to take the effects of injection locking into account is to measure the effective EFC gain with the loop closed. What I use for a finial calibration check is the 2G turn over, which I measure very accurately by other means before hand and then use that as a known freq offset to check operation and calibration. Of course there are any number of other ways. As far as temperature having ANY effect on EFC gain, that is a total NON issue. If temperature had any effect on EFC Gain then Temperature would also effect Osc Frequency at a fixed EFC voltage, which would then effect the OSC freq drift and stability, that would then effect anything that the Osc was used for, NOT just the TPLL. The TPLL actually has a slight advantage over other methods, because the PLL will adjust the freq to be correct, even if the EFC effect should change. I think it is reasonable to assume that a TPLL weighs in at about 200 USD with all support mixers, amplifiers, ADCs etc. if you don't have the parts It is still a fairly cheap solution. Yes I think that is ONE reasonable number to use and a fair conclusion. BUT there are others. The EBAY cost of the TPLL can be easy under $10, not including the reference Osc and the ADC. Do note, NONE of items above are plural, Only one is needed per system unlike some other methods. Because the cost of the Ref Osc is so variable and depends so much on what one is doing, I have noticed that its cost is generally not included in the base price. I think even on the $20K+ TSC 5120A that the reference Osc is an extra cost option. The ADC is another BIG variable, depending on your needs and skill level and junk box, almost no limit in cost at the high end, and can be as low as $0.00 dollars if you are a student doing a science project. It can also be as low as $1.00 if one is good at programming PICS or other micros with built in ADC's. The only other major part in the TPLL with any cost over $1 is the Phase detector. The one I use most is a micro-circuits $15 single price device, but I've used all sorts of dual balanced mixers, and if one is real cheap and good at design, I have found that a PD based on a 50 cent XOR gate works fine. ws * Bruce On 14 June 2010 10:46, Magnus Danielson Posted: Steve Still puts it in the mid-tau range as a method. The useful range and precision of a particular implementation of the method will vary. By putting a GPSDO in the usual place of the DUT and putting the 10811 in place of the reference oscillator it could work well beyond the 1000s point. CAVEAT: this only works for a DUT that has an EFC that is reasonably linear. EFC linearity will remain an issue for analog oscillators. The oscillator gain will differ depending on offset voltage and temperature. So if you are just thinking about the TPLL for taking ADEV data from 0.1 to 1000 sec, then you're are missing 90% of the other useful stuff it can do as good or better than most anything thing else out there, and all for the same $10 (my cost). The typical price-tag of a 10811 is in 100-150 USD. I think it is reasonable to assume that a TPLL weighs in at about 200 USD with all support mixers, amplifiers, ADCs etc.
Re: [time-nuts] Advantages Disadvantages of the TPLL Method
WarrenS wrote: Bruce posted If and only if injection locking isn't significant. No problem then, because it is not significant. For each and every oscillator pair someone may try? Can place this one under the 'ADVANTAGE' side. That's descending into the murky realms of pseudoscience. At best you've only shown this to be true for the particular oscillator pair being compared. Not only must the effect of injection locking be insignificant for the reference, it has to be insignificant for the test oscillator as well. If injection locking is an issue the efc gain with the loop open will differ from the efc gain with the loop closed. I have tested this thoroughly in many ways. I do understand the concerns and doubts, especially with an unbuffered HP 10811 as the reference. The 10811s are pretty sensitive to injection locking and phase pulling. Unlike most other methods, one of the many unique properties that the TPLL method has is that injection locking is normally not a problem with it. It will change the loop parameters in particular the efc gain. Its just a matter of how much it affects the efc gain. I find it is generally unnecessary to buffer either the Ref Osc or the DUT. This is one of the many features that helps make the simple TPLL so simple. (also it does not hurt or change anything to add a proper buffer) The lack of injection locking is one of the advantages that contributes to its exceptional and unbelievable performance. But Adler's equation indicates that an oscillator is much more to susceptible to injection effects when the injected signal frequency is very close to the oscillator frequency. I did not leave the buffers out of the simple TPLL BB that was tested because of my lack of knowledge, but because of my extra knowledge on the subject that showed that they were unnecessary. More than once, I have tried to explain the reason why injection locking is not a problem with my version of the TPLL method, but until one proves it for their self, more words from me will not help. I do understand the skepticism and doubt, and I know why it is so hard to believe this for those that have not worked with is this type of method before. I guess someone should write one of those fancy math papers, if it has not already been done, that explains it in more convincing terms than I've been able to. It is hard for me to believe that paper has not already been written, But then it is hard for me to believe that the TPLL is not used more often. There are plenty of places that one of the TPLL methods well give the best overall solution. ws *** Bruce [time-nuts] Advantages Disadvantages of the TPLL Method Bruce Griffiths bruce.griffiths at xtra.co.nz WarrenS wrote: Long explanations, cause I try to explain, the best I can, when I say something is WRONG or misleading Magnus Posted: EFC linearity will remain an issue for analog oscillators. The oscillator gain will differ depending on offset voltage and temperature. TRUE it is an issue, but somewhat misleading because it need NOT be a problem or limitation (mostly) EFC Linearity can be an issue because the TPLL is limited by the performance of the reference oscillator in lots of ways. BUT Oscillator EFC gain or linearity are not likely to be of much concern or a limitation for high end performance. The gain nonlinearity I've measured can vary two to one over the full range of a good Oscillator but it is more like 10% over the normally used range, if one stays well away from the end points. NOT so good but livable if you are not making something real accurate. BUT For all my accurate stuff, when using a HP 10811, I limit the full-scale change to 1e-9 or 1e-8 at most. This uses such a small part of the total EFC range, that the nonlinearity effects are generally below the noise level and of little concern at all. The fact that Oscillator gain does differ with the EFC voltage (offset voltage), means if you want to get max accuracy out of the TPLL, it will need to be calibrated at the EFC offset voltage it is being used at. One simple solution, if the OSC also has a independent manual Freq adjustment like the single oven 10811, is to use it always set the EFC voltage to be near zero volts. BTW calibration need not be much of a problem, because it can be a static calibration. If and only if injection locking isn't significant. This needs to be established for each setup. The simplest way to take the effects of injection locking into account is to measure the effective EFC gain with the loop closed. What I use for a finial calibration check is the 2G turn over, which I measure very accurately by other means before hand and then use that as a known freq offset to check operation and calibration. Of course there are any number of other ways. As far as temperature having ANY effect on EFC gain, that is a total NON issue. If temperature had any effect on EFC Gain then Temperature
Re: [time-nuts] Advantages Disadvantages of the TPLL Method
WarrenS wrote: subject: Advantages Disadvantages of the TPLL Method. Here is a new and unique Idea that may be useful for many. Rather than focusing on what some members may or may not already know, or how good or bad one specific working BB configuration is. How about focusing on what the TPLL method can and can not do well. If someone will make a place to post and compile a couple of list, I can start it off with what I've learned so far: DISADVANTAGES of the TPLL method: --- #1) The TPLL method is limited by it's reference OSC. This isn't necessarily correct, one could use a pair of tight PLL loops and use correlation techniques to reduce the contribution of the reference oscillator noise. The ref osc (or the DUT) needs to have an Analog/or Digital EFC control input with a bandwidth that is wider than the desired Tau0 #2) It basically measures Freq and not Phase differences, and few understand how and why it works so well or it's many advantages. This is not true, there is no inherent SNR advantage in measuring frequency changes as opposed to measuring phase differences. When the phase measurement system and the frequency measurement systems being compared have the same noise bandwidth then the measurement floors are comparable. For example, the TSC5120A is a narrow band system based on measuring phase differences with a comparable or lower noise floor than your implementation of the tight PLL. The common technique of using a time interval counter to measure the phase difference between 2 RF signals once ever second or so is a wideband technique with severe undersampling, consequently the system noise floor is much higher than for narrow bandwidth techniques. If the phase difference between the 2 signals were measured more frequently and digitally low pass filtered the noise will be much lower. Since one has to calculate average frequency from the frequency samples by integration/averaging this is mathematically equivalent to reconstructing the phase change between the start and end of the averaging time (Tau0). One effect of undersampling is to convert (in the sampled data) a proportion of any flicker phase noise (and other non white phase noise components) to white phase noise. The effect of this is to change the ADEV vs Tau plots from their true shape. With a single pole RC filter the required minimum sampling rate to ensure that such effects are acceptably small cannot be known unless the phase noise spectra of the 2 oscillators being compared is known. However the extra phase noise filtering due to the finite PLL bandwidth (including any EFC filtering built in to the reference oscillator) allows an estimate of the maximum sampling rate likely to be required to ensure that such phase noise whitening effects are acceptably small. #3) TBD ADVANTAGES of the TPLL method: --- 1 thru 30) same as I've posted several times before. I'm sure others will find many more if they try it or at least understand it better. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Advantages Disadvantages of the TPLL Method
Magnus Danielson wrote: On 06/12/2010 11:29 PM, Bruce Griffiths wrote: WarrenS wrote: subject: Advantages Disadvantages of the TPLL Method. Here is a new and unique Idea that may be useful for many. Rather than focusing on what some members may or may not already know, or how good or bad one specific working BB configuration is. How about focusing on what the TPLL method can and can not do well. If someone will make a place to post and compile a couple of list, I can start it off with what I've learned so far: DISADVANTAGES of the TPLL method: --- #1) The TPLL method is limited by it's reference OSC. This isn't necessarily correct, one could use a pair of tight PLL loops and use correlation techniques to reduce the contribution of the reference oscillator noise. True. The same technique is being used for LPLL phase noise measurements. The reference oscillator will still be a limit, but wither you can go below the reference oscillator noise or not is what makes the difference. Such a setup costs about twice of a single-channel TPLL. Usually there is two ADC channels available. Yes the cost of the reference oscillator dominates the system cost, the additional $10 (omitting the cost of the phase detector) to implement the tight PLL is relatively insignificant. The cross-correlation processing isn't too hard to achieve and is efficiently performed using FFTs and a little support-processing. FFTW is a good tool to toss the FFT processing to. The remaining wrapping is in a few ten lines of codes or so. Going down the FFT path will give the frequency plot for free, getting it back into the time-domain cost extra. If one is calculating the FFT then it is possible to calculate ADEV directly from the FFT (of the frequency samples) with little additional effort, for the relevant formulae see: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf Note such processing doesn't increase the cost of the system as one needs a PC to calculate frequency stability measures, unless one wants/needs to do it in real time. One disadvantage of a tight PLL system is that finite EFC range and EFC non linearity may preclude its application to noisier sources. Linearising the EFC transfer function will help but the reference oscillator EFC range will ultimately provide an upper limit to the measurable noise. The ref osc (or the DUT) needs to have an Analog/or Digital EFC control input with a bandwidth that is wider than the desired Tau0 #2) It basically measures Freq and not Phase differences, and few understand how and why it works so well or it's many advantages. This is not true, there is no inherent SNR advantage in measuring frequency changes as opposed to measuring phase differences. When the phase measurement system and the frequency measurement systems being compared have the same noise bandwidth then the measurement floors are comparable. For example, the TSC5120A is a narrow band system based on measuring phase differences with a comparable or lower noise floor than your implementation of the tight PLL. The common technique of using a time interval counter to measure the phase difference between 2 RF signals once ever second or so is a wideband technique with severe undersampling, consequently the system noise floor is much higher than for narrow bandwidth techniques. If the phase difference between the 2 signals were measured more frequently and digitally low pass filtered the noise will be much lower. Using time-stamping counters at high rate would be possible if being able to cope with the rate of samples. You want a frontend to do that if you want to run continously. As for digital filtering. When doing measurements in the 0,1 - 1000 s range for the G.813 measurements, a 10 Hz low-pass filter is being required. Since one has to calculate average frequency from the frequency samples by integration/averaging this is mathematically equivalent to reconstructing the phase change between the start and end of the averaging time (Tau0). Depends on the details. Some counters (SR620 for instance) can have biases for frequency data which their time-difference measures do not have. A TPLL does not suffer from that particular problem, as it cranks out its frequency estimation by a different method. Yes, but I thought that we were calculating the required averages from the frequency (EFC) samples by approximating the required integrals. One effect of undersampling is to convert (in the sampled data) a proportion of any flicker phase noise (and other non white phase noise components) to white phase noise. The effect of this is to change the ADEV vs Tau plots from their true shape. Care to hand a reference or two for this statement? References for the whitening effect of undersampling: http://www.obs-besancon.fr/tf/publis/metrologia98a.pdf http://www.obs-besancon.fr/tf/publis/metrologia98b.pdf
Re: [time-nuts] Advantages Disadvantages of the TPLL Method
Another disadvantage of the Tight PLL system that only applies to multichannel systems is that a dedicated reference oscillator is required for each channel. i.e. for an N channel system N reference oscillators are required. If correlation techniques were to be employed then an N channel system requires 2N reference oscillators. N channel versions of Dual Mixer systems by contrast only need a single offset oscillator and a single reference oscillator. Similarly an N channel heterodyne system only requires a single offset oscillator. An N channel direct RF phase sampling system (like that employed by the 2 channel TSC5120A) only requires a single samplign clock source. An N channel time interval counter that periodically (eg at a 1Hz rate) measures phase differences between 2 RF signals only requires a single reference source. The above system can be regarded as an undersampled version of the direct RF phase sampling system. The poor cost scaling of the tight PLL system is another reason why it has fallen out of favour for those who have more than 2 frequency standards to compare simultaneously. Bruce WarrenS wrote: Great start Now if we just had a list that someone would add the advantages and disadvantages to, so that any non relevant stuff could be easily seen and removed or moved to a third list, It would all become much clearer. ws Magnus Danielson wrote: On 06/12/2010 11:29 PM, Bruce Griffiths wrote: WarrenS wrote: subject: Advantages Disadvantages of the TPLL Method. Here is a new and unique Idea that may be useful for many. Rather than focusing on what some members may or may not already know, or how good or bad one specific working BB configuration is. How about focusing on what the TPLL method can and can not do well. If someone will make a place to post and compile a couple of list, I can start it off with what I've learned so far: DISADVANTAGES of the TPLL method: --- #1) The TPLL method is limited by it's reference OSC. This isn't necessarily correct, one could use a pair of tight PLL loops and use correlation techniques to reduce the contribution of the reference oscillator noise. True. The same technique is being used for LPLL phase noise measurements. The reference oscillator will still be a limit, but wither you can go below the reference oscillator noise or not is what makes the difference. Such a setup costs about twice of a single-channel TPLL. Usually there is two ADC channels available. Yes the cost of the reference oscillator dominates the system cost, the additional $10 (omitting the cost of the phase detector) to implement the tight PLL is relatively insignificant. The cross-correlation processing isn't too hard to achieve and is efficiently performed using FFTs and a little support-processing. FFTW is a good tool to toss the FFT processing to. The remaining wrapping is in a few ten lines of codes or so. Going down the FFT path will give the frequency plot for free, getting it back into the time-domain cost extra. If one is calculating the FFT then it is possible to calculate ADEV directly from the FFT (of the frequency samples) with little additional effort, for the relevant formulae see: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf Note such processing doesn't increase the cost of the system as one needs a PC to calculate frequency stability measures, unless one wants/needs to do it in real time. One disadvantage of a tight PLL system is that finite EFC range and EFC non linearity may preclude its application to noisier sources. Linearising the EFC transfer function will help but the reference oscillator EFC range will ultimately provide an upper limit to the measurable noise. The ref osc (or the DUT) needs to have an Analog/or Digital EFC control input with a bandwidth that is wider than the desired Tau0 #2) It basically measures Freq and not Phase differences, and few understand how and why it works so well or it's many advantages. This is not true, there is no inherent SNR advantage in measuring frequency changes as opposed to measuring phase differences. When the phase measurement system and the frequency measurement systems being compared have the same noise bandwidth then the measurement floors are comparable. For example, the TSC5120A is a narrow band system based on measuring phase differences with a comparable or lower noise floor than your implementation of the tight PLL. The common technique of using a time interval counter to measure the phase difference between 2 RF signals once ever second or so is a wideband technique with severe undersampling, consequently the system noise floor is much higher than for narrow bandwidth techniques. If the phase difference between the 2 signals were measured more frequently and digitally low pass filtered the noise will be much lower. Using time-stamping counters at high rate
Re: [time-nuts] Advantages Disadvantages of the TPLL Method
WarrenS wrote: Thanks for the positive contrbution, A good example of one of the TPLL's obvious disadvantages. The simple cheap analog version of the TPLL is limited by it's need to have a dedicated Ref OSC. One way I have got around that problem, which would not apply to all, is to put the DUT unit as the controlled OSC, and use a special Tbolt as the reference Oscillator. The other way around the problem is the Digital version of the TPLL that uses DSS. BTW that limitation is not nearly as big as one would think. This is because the long term accuracy is already limited by the reference osc, so one would not generally use this kind of system out past 1000 sec or so anyway. So If doing long term multichannel Osc, One would likely be MUCH better off with a more basic undersampled Phase system for long term testing and just go thru and cheek each Osc one at time for a short time with a low tau tester such as this type. Keep the advantages and disadvantages coming in, so the Time Nuts can compare which methods work best for their application. Now if we just had some place to log the responses. Summery: If you have multi oscillators to test simultaneously that do not have EFC input, and that you want to do continuous sampling on, and do not have multiple TSC boxes, the TPLL is not the right tool for the job. Be better off with one simple lower resolution multiplexed time stamped TI phase system and a single TPLL. If one has a production requirement to test/compare several hundred oscillators simultaneously, the TSC5120A and its variants, being 2 channel instruments, aren't really that useful even if one could afford several hundred of them. Such a requirement may be difficult to meet within a modest budget whilst still achieving the performance requirements (eg 1E-13/tau system noise). Even with a much smaller number of oscillators (eg 8 -16) devising an affordable measurement system may be challenging. Bruce Bruce posted: The poor cost scaling of the tight PLL system is another reason why it has fallen out of favour for those who have more than 2 frequency standards to compare simultaneously. Thanks for that opinion, but I don't think we should list the above as a unique disadvantage. Maybe need a new column heading for that one, Any name suggestions? Does not sound all that valid or unique of a reason to me. It seems the same can be said about a TSC or any new high cost system. I would think a more important reason is that the simple TPLL is not a universal do all system. Because the simple analog version is Limited by it's reference Osc in many ways, This does give it some possible major disadvantages like not working so good with a CS or Rb standard. If one has more time than money, there are ways around that. ws *** Bruce Griffiths bruce.griffiths at xtra.co.nz Sun Jun 13 01:25:13 UTC 2010 Another disadvantage of the Tight PLL system that only applies to multichannel systems is that a dedicated reference oscillator is required for each channel. i.e. for an N channel system N reference oscillators are required. If correlation techniques were to be employed then an N channel system requires 2N reference oscillators. N channel versions of Dual Mixer systems by contrast only need a single offset oscillator and a single reference oscillator. Similarly an N channel heterodyne system only requires a single offset oscillator. An N channel direct RF phase sampling system (like that employed by the 2 channel TSC5120A) only requires a single samplign clock source. An N channel time interval counter that periodically (eg at a 1Hz rate) measures phase differences between 2 RF signals only requires a single reference source. The above system can be regarded as an undersampled version of the direct RF phase sampling system. The poor cost scaling of the tight PLL system is another reason why it has fallen out of favour for those who have more than 2 frequency standards to compare simultaneously. Bruce WarrenS wrote: Great start Now if we just had a list that someone would add the advantages and disadvantages to, so that any non relevant stuff could be easily seen and removed or moved to a third list, It would all become much clearer. ws Magnus Danielson wrote: On 06/12/2010 11:29 PM, Bruce Griffiths wrote: WarrenS wrote: subject: Advantages Disadvantages of the TPLL Method. Here is a new and unique Idea that may be useful for many. Rather than focusing on what some members may or may not already know, or how good or bad one specific working BB configuration is. How about focusing on what the TPLL method can and can not do well. If someone will make a place to post and compile a couple of list, I can start it off with what I've learned so far: DISADVANTAGES of the TPLL method: --- #1) The TPLL method is limited by it's reference OSC. This isn't necessarily
Re: [time-nuts] UTC and leap seconds
Tom Van Baak wrote: Beside the general theoretical considerations as of what answer is more acceptable (sincerely I agree so far) and what method could be used to solve the matter, can anybody out there point me please to any article on actual measurements of the variation rate of the earth's rotational speed, not based on clocks? Antonio, Consider that you need at least two clocks before you can make a rate measurement. One is the DUT; the other the REF. So it is not possible to measure the earth (DUT) without using some other clock (REF). Make sense? (Speculative hint: We accept that the universe is expanding. Might this affect the fine structure of matter, including cesium atoms? Is there any adverse proof? What is easier to think? a) the expansion of the universe doesn't affect at all the properties of matter. b) it might.). There is no small amount of effort being put into this question. The results are not usually given as yes/no, zero or non-zero. Instead they just calmly establish a new lower bound on what the drift rate might be. Whether the answer is (a) or (b) doesn't change the fact that the earth day is a poor clock compared with other clocks now available. Besides tidal friction effects which might be hard to imagine, or lunar effects which you already know about, note that every time it rains or glaciers form and melt it changes the angular momentum of the poor spinning planet. Surely you mean that it changes the moment of inertia of the planet??? Then again, many OCXO are also affected by humidity... /tvb Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Tight-PLL - YOU DON'T NEED TO READ IT IF YOUR FED-UP WITH THE THREAD SO HIT DELETE NOW!
Wrong again. The integration/averaging referred to occurs when one counts the output transitions of the VFC for a fixed time interval. This process needs to be replicated using the sampled EFC data if one is to measure ADEV. Bruce Steve Rooke wrote: I think I have found the source of the integration issue. I've spent some considerable time ploughing through as many sources of descriptions on ADEV, AVAR and the tight-PLL method. I've even tried looking for the infamous finite time interval integrator which seems to be highly notable by it's complete absence on Google. Well, eventually the answer struck me directly in the eye, the source of the integrate issue comes directly down to the original paper that Warren posted a link for:- D. Tight phase lock loop method The second type of phase lock loop method (shown in figure 1.7) is essentially the same as the first in figure 1.6 except that in this case the loop is in a tight phase lock condition; i.e., the response time of the loop is much shorter than the sample times of interest--typically a few milliseconds. In such a case, the phase fluctuations are being integrated so that the voltage output is proportional to the frequency fluctuations between the two oscillators and is no longer proportional to the phase fluctuations (for sample times longer than the response time of the loop). A bias box is used to adjust the voltage on the varicap to a tuning point that is fairly linear and of a reasonable value. The voltage fluctuations prior to the bias box (biased slightly away from zero) may be fed to a voltage to frequency converter which in turn is fed to a frequency counter where one may read out the frequency fluctuations with great amplification of the instabilities between this pair of oscillators. The frequency counter data are logged with a data logging device. The coefficient of the varicap and the coefficient of the voltage to frequency converter are used to determine the fractional frequency fluctuations, yi, between the oscillators, where i denotes the ith measurement as shown in figure 1.7. It is not difficult to achieve a sensitivity of a part in 1014 per Hz resolution of the frequency counter, so one has excellent precision capabilities with this system. http://tf.nist.gov/phase/Properties/one.htm The relevant section here is the response time of the loop is much shorter than the sample times of interest--typically a few milliseconds. In such a case, the phase fluctuations are being integrated so that the voltage output is proportional to the frequency fluctuations. So what this says is that by incorporating a PLL-loop filter that has a B/W much wider than the sample time, the phase fluctuations are integrated into the reference oscillator such that the control voltage of the tight-PLL now reads frequency which is unlike the loose-PLL which directly records the phase relationship between the oscillators. So the term integrated here is used a verb and not a noun, therefore it is an intrinsic function of the design not a separate process. Steve -- Steve Rooke - ZL3TUV G8KVD The only reason for time is so that everything doesn't happen at once. - Einstein ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Small DMTD System
Richard H McCorkle wrote: Time-Nuts, There has been much discussion on this list about methods of measuring short-term stability. I wanted to make the list aware of a new paper describing a small DMTD system. The system was developed by William Riley, author of STABLE32, and is described in detail with schematics and test results at: http://www.wriley.com/A%20Small%20DMTD%20System.pdf Richard Its amazing what one can do with a ZCD design that totally ignores the principles of low noise design. With a little redesign and addition of a few inexpensive parts it should be capable of much better performance. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Tight-PLL - YOU DON'T NEED TO READ IT IF YOUR FED-UP WITH THE THREAD SO HIT DELETE NOW!
Steve Rooke wrote: On 5 June 2010 19:07, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Wrong again. No, I'm not wrong Bruce. Your contribution is largely irrelevant to the original discussion. The effect of the PLL itself is (or should be) well understood. However various assertions about the minimum usable value of Tau take no account of the low pass filtering built into the 10811 EFC circuit. The 100k series resistors plus the capacitance of the EFC varicap (50-100pF??) will limit the minimum usable value of Tau. The integration/averaging referred to occurs when one counts the output transitions of the VFC for a fixed time interval. This process needs to be replicated using the sampled EFC data if one is to measure ADEV. This process is exactly replicated by oversampling the EFC and determining the average for a fixed time period. A various times Warren has both claimed to do this and at others appears to deny it. A clear description of the details of the actual signal processing used is sadly lacking. If and only if the average is calculated sufficiently accurately. Using a rectangular approximation with sampled data may not be as accurate as one may expect. It never ceases to amaze me why the well established and more accurate methods known aren't used (details are all given in the paper I cited), all it requires is a suitable program running on a PC. The correct processing should have no effect on the hardware cost. The $10 cost is also misleading as the mixers aren't free nor is the 10811 or its equivalent. The assertion that this technique is new seems to be somewhat dubious as it appears to have been known for several decades. If you can't see that this performs exactly the same function, I don't know what will convince you. Steve Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Z3816a
Robert Benward wrote: Joe, Thank you for your inputs! The Z3815A I suspect is low voltage If and only if a ~24-48V supply is considered low voltage. The 10MHz outputs level is only around +4dBm though there are 4 10MHz outputs. , it has a large PC card/motherboard type of connector with a large molex/amphenol power connector. There appears to be a pair of DC-DC modules inside. You would need to make some internal taps to get the 10MHz out to a BNC. I also read somewhere an author's disappointment on the performance of the (his?) E1938A compare to the 10811. I have one that is stand alone, running continuously. I was under the impression that the maximum number of satellites in view is about 12 (24 active in a constellation hemisphere view), the other channels I think are used for WAAS and other data channels. Bob Bruce - Original Message - From: J. L. Trantham jlt...@att.net To: 'Discussion of precise time and frequency measurement' time-nuts@febo.com Sent: Friday, June 04, 2010 10:56 PM Subject: Re: [time-nuts] Z3816a Correction. Z3805A tracks 16 satellites. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of J. L. Trantham Sent: Friday, June 04, 2010 9:51 PM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] Z3816a I have Z3816A's and TBolt's. Both do a great job. For the TBolt, I use TBoltMon.exe to monitor and for the Z3816A, I use SatStat. The Z3816A requires a 'null modem' adapter or an appropriately wired serial cable to connect to the serial port of the computer where as the TBolt uses a straight serial cable. The only potential advantage of the Z3815A that I know of is the fact that it uses the E1938A oscillator. I have a couple of these oscillators and they are superior in performance. However, I have no firsthand knowledge of performance of the Z3816A versus the Z3815A. The Z3816A and the TBolt tracks up to 8 satellites. The Z3816A also tells you what other satellites are under consideration but are not being tracked, perhaps up to 8 other non-used satellites. I have at most seen the Z3816A refer to a total of 11 satellites (8 tracked and 3 not tracked) but my antenna is not in an ideal location. I do not know about the Z3815A. The Z3805A tracks up to 12 satellites, if I recall correctly, and probably should also be included in the comparison list if you are looking for a GPSDO. The Z3816A comes in at least two variants, one powered by a DC supply (I use 30 VDC for mine) and the other powered by 120 VAC. I don't know about the Z3815A. Good luck, Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert Benward Sent: Friday, June 04, 2010 7:13 PM To: time-nuts Subject: [time-nuts] Z3816a Hi All, Is there any downside to a model Z3816A? I also see a Z3815A on Ebay, but the connector arrangement is not attractive. Does anyone have a link to a comparison chart between all these models? ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
Steve Rooke wrote: On 4 June 2010 08:32, Charles P. Steinmetz charles_steinm...@lavabit.com wrote: If I may be allowed to summarize, it appears that Warren and Bruce agree that integration is necessary to produce true ADEV results. Warren asserts that the low-pass filtering his method uses is close enough to integration to provide a useful approximation to ADEV, while Bruce disagrees. So, the remaining points of contention seem to be: 1. How close can a LPF implementation come to integration in ADEV calculations, and Well, Warren uses two stages of integration. There has already been talk of the simple R/C filter in the feedback loop. Unless my education in electronics was completely wrong, the series R/C circuit forms a simple LPF and is an integrator (assuming that the resistor is in series with the input and the capacitor is in parallel with the output). See http://en.wikipedia.org/wiki/Integrator_circuit, http://en.wikipedia.org/wiki/RC_filter. Sorry these are not academic papers but if you spot something wrong please feel free to edit them appropriately. This first stage of integration is set at a much wider frequency than tau0 and forms the PLL-loop filter allowing it to track the FAST changes of a noisy unknown oscillator. That last bit is very important and something some previous attempts at this method failed to resolve. A cascaded low pass filter and and a finite time interval integrator are required. A single RC LP filter can't approximate this. Its either a low pass filter or a crude approximation to an integrator not both. Now there is a noisy control voltage on the reference oscillator and it is absolutely no good trying to make a single measurement at tau0 because the settling time of the filter has not been constrained to it so it will not give an integrated mean value. This where the second stage of integration comes in which is the oversampling which takes a number of readings during tau0 (please correct me if I have the terminology wrong here) which are then averaged to give a mean, integrated, value of the control voltage for tau0. Speculative nonsense sampling by itself integrates nothing unless one uses an integrator to do the sampling. Even when the finite bandwidth of the sampler is taken into account the equivalent averaging time will be too short and not under user control. However the samples (if the sampling rate is sufficiently large) contain sufficient information for the required finite time integrator output values (or frequency averages) to be calculated. A simple rectangular integration approximation may not be sufficient in all cases. The sampling process actually tends to whiten the sampled phase noise spectrum. The amount of false white phase noise contributed by the sampling decreases as the sampling rate increases. A simple RC low pass filter may not be a a particularly good choice in this regard. So why two stages, look closely above, until the idea of oversampling was tried, the PLL-loop filter had to have a settling time, IE. cutoff frequency, equal to tau0 so that the measurement at tau0 reflected the mean, average, integrated, value for that tau0 period. But if a filter with that sort of cutoff is used then the reference oscillator is not able to track noise on the unknown oscillator at all and it would give results for things like flicker noise, random walk, etc, which were lower than the actual values. Now have a look at the top end of John's graphs where there is a divergence. The divergence at the top end of the graphs should be treated with extreme caution one needs to know the size of the associated error bars to be able to make statistically meaningful conclusions. In general the error bars tend to be large in this region. 2. How close to true ADEV is good enough? well, considering we have integrated frequency measurements at tau0 intervals, there is little wonder that it correlates closely to ADEV because that's exactly what it is. This cannot be so for each and every signal source if the weighting function (equivalent filter) doesn't closely match that used in the definition of AVAR. Without the integration/averaging the equivalent filter closely match the required filter at all frequencies. I humbly submit that trading insults has become too dreary for words, and that neither Warren nor Bruce will ever convince the other on the latter point. Well, I've been on this list long enough to know that Bruce will always resort to that sort of behaviour when he is boxed into a corner or cannot get his point of view accepted. Anyone who speaks up against him is usually put in their place. This saga has come about because someone dared to challenge him so we have been subjected to his tantrums. The saga originated because of the wildly inaccurate claims and very woolly explanation as to what signal processing was used. A few equations and a circuit diagram or 2 would have
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
WarrenS wrote: Bruce posted The RC filter doesn't accurately integrate the frequency difference over time interval Tau0. For you to even state that means you still have NO idea what I'm doing, It is getting sort of sad. Correct the RC filter is not an integrator, it is used for the combination Bandwidth and anti-aliasing filter. It is the oversampling average that does the integration. How? Rectangular integration isnt particularly accurate or efficient, better techniques exist. What would explain a lot is, if you do not know what oversampling even is? Try to desist from the pathetic attempts at insults as they merely distract from the real questions about the signal processing techniques adopted. You need to get yourself a refresher course on the advantages of oversampling to do integration, brick wall filtering, anti-aliasing and why a single RC works just fine for integration when oversampling is used and why you don't need anything but simple averaging of sum n samples /n when oversampling is used. Don't need all the unnecessary fir filter crap, just oversample. Not so, as anyone with a comprehensive understanding of the subject will attest. Seat of the pants methods produce misleading predictions when noise isnt statistically stationary. If you have spare bandwidth like I have, then it sure saves a lot of stuff. Ever hear of KISS'. Most are aware of the principle but over simplification leads to erroneous results. You need to ask someone to explain that to you some day, along with close enough Hint, the simple Tester BB only takes ONE IC and it is just a single op Amp. That performance metric is irrelevant if it doesnt measure the desired quantity for all cases of interest. NB the case spectrum will vary from one user to another so the limitations of the technique need to be well known. These limitations will include limits on the phase noise spectra of the devices being compared. And Although John's Software makes it all much more user friendly and makes user mistakes less likely to occur, It is not needed. Works just fine with no special S/W code or filter S/W. AND it still does integration just fine. (Send me that data file if you want to see how it works). You seem to be unaware of just how easy it is to create a dataset for which any given algorithm will fail catastrophically. ws ** Bruce last posted: John Miles wrote: The integration secret (which is no secret to anyone but Bruce) is to analog filter, Oversample, then average the Frequency data at a rate much faster than the tau0 data rate. Which again is misleading as you specify neither the averaging method nor the analog filter. I can't speak for the analog side as I never saw a schematic of the PLL, but it may be worthwhile to point out that the averaging code in question is in SOURCE_DI154_proc() in ti.cpp, which is installed with http://www.ke5fx.com/gpib/setup.exe . This is my code, not Warren's. It does a simple boxcar average on phase-difference data, the same as my TSC 5120 acquisition routine does. Previous tests indicated that simple averaging yields a good match to most ADEV graphs on TSC's LCD display, so I used it for the PLL DAQ code as well. I also tried a Kaiser-synthesized FIR kernel for decimating the incoming TSC data, but found that its conformance against the TSC's display was worse than what I saw with the simple average. More work needs to be done here. When will you understand that phase differences and differences of average frequency (unit weight to frequency measures over the sampling interval zero weight outside) are equivalent. One subtlety is the question of whether to average (or otherwise filter) the DAQ voltage readings immediately after they're acquired and linearly scaled to frequency-difference values, versus after conversion of the frequency-difference values to phase differences. I found that the best agreement with the TSC plots was obtained by doing the latter: val = (read and scale the DAQ voltage) // val is now a frequency difference // averaging val here yields somewhat higher // sigma(tau) values in the first few bins // after tau0 val = last_phase + (val / DI154_RATE_HZ); last_phase = val; This appears to use a rectangular approximation to the required integral. A trapezoidal or even Simpson's rule integration technique should be more accurate for a given sample rate. One could even try a higher order polynomial fit to the sample points, however this isnt the optimum technique to use. If one uses WKS interpolation to reconstruct the continuous frequency vs time function and integrates the result for a finite time interval (Tau0) then one ends up with a digital filter with infinite number of terms. Since an infinite number of samples is required to do this using a suitable window function is probably advisable. The paper (below) illustrates how AVAR etc can be calculated from the sampled
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
Steve Rooke wrote: On 3 June 2010 15:46, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: WarrenS wrote: As Bruce says It remains a mystery to him why this works. It doesnt, it only appears to in a very restricted set of circumstances. Bruce, I don't understand you, when presented with visual evidence that this method works you still deny it. What visual evidence?? There is no proof that the technique works well in every case. Only that for the range of Tau tested and for the particular source pair used that it appears to. Not one of my best skills, but I'll try to explain it once again. Now that they see it works, maybe someone else will be able to put this into words that Bruce will be able to finally understand. The only requirement needed for the Frequency data log to be give correct ADEV readings, is to get good, Averaged, integrated, Frequency data, with no dead time, and no aliasing, over the tau0 time period. Each Tau0 Frequency sample is ideally completely independent from all others. If it can do one right then it can get them ALL right. In a single tau0 sample there is NO SUCH THING as a certain type of long term noise, Just the average freq over that single time period. Misleading as usual, your knowledge of statistics is woefully inadequate leading to incorrect conclusions. Well, what are are the woefully inadequate conclusions then? Please give us your full reasoning. A simple example is that for a small number of samples a stability metric like the ordinary (unfiltered) phase variance standard deviation may appear to be stable, whereas with a sufficiently large number of samples the instability of the metric itself becomes evident whenever divergent noise processes like flicker phase noise, random walk frequency noise are present. /Each Tau0 Frequency sample is ideally completely independent from all others. / The above statement is incorrect as the finite bandwidth necessarily imparts a correlation between samples they can only be strictly independent if the bandwidth is infinite. /In a single tau0 sample there is NO SUCH THING as a certain type of long term noise, Just the average freq over that single time period. / The above statement imparts no useful information. It would be much easier and less bandwidth wasted if the circuit schematics and useful documentation on the algorithms employed were available. Extracting any useful information seems somewhat akin to pulling teeth. The crucial integration/averaging to get good tau0 data, that Bruce can not see for some unknown reason, is done Only in your imagination. One would assume that this method only works when Warren does it as his imagination is required for it to work, but wait, John Miles has managed to get point for point identical data against a TSC, how can that be Bruce? Please give answers, not insults. Read the following paper: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf which shows the relationship between AVAR etc, filters and the ordinary phase variance. The paper also outlines the techniques that should be used with the sampled frequency difference data from a tight PLL. with an analog filter set to about the Tau0 Freq and by oversampling at about about a 10 to one ratio, and averaging the oversampled frequency readings down to tau0. That doesn't work as it has the wrong transfer function. Again, it it does not work, how come the evidence shows that it does, how do you explain that Bruce? The evidence doesn't show this at all. It merely indicates that for the devices tested that the phase noise spectral components in the region where the filter responses of the ADEV and WDEV differ (its not ADEV so it shouldnt be labelled as such) dont appear to be significant for the 2 sources compared and the tau range over which the testing was done. Extrapolation of such results to predict that the technique will produce such agreement with other devices with differing phase noise characteristics is unrelaible. You are confusing producing the same numbers in specific cases with the ability to do so in general. There is no guarantee that such agreement will occur with a given pair of sources. Such agreement in general isnt possible as the equivalent phase noise filters have different frequency responses. Stability measures like AVAR can be shown to be the equal to the ordinary variance of the phase difference at the output of a very specific phase noise filter. WDEV has a phase noise filter with a different frequency response so that it doesnt actually measure ADEV. It is not perfect, but plenty close enough for the plot to match the output of the TSC 5120A over the whole tau range. There are a few other subtle details on how to insure that aliasing and over filtering do not become a problem, but first things first, one needs to understand how the integration is
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
WarrenS wrote: Bruce posted: It would be much easier if Warren limited his commentary to the actual results and omitted the wild speculation OK, It works with everything that it has been tried on and gives the same answers as the TSC5120A, including the Osc shake test. Now your turn to try and find something that it does not work on. ( see end Item) Phase is the integral of frequency so phase differences sampled at intervals of say T are equivalent to frequencies averaged over time T and sampled at the end of the sample interval. Thus sampling the time average frequency every T seconds is equivalent to sampling the phase difference every T seconds. Equivalent Information Yes, but not equal to at all. Please tell me you think you can now process both of those data set with the same algorithm and get the same results? Then we can all have something to laugh at. Another somewhat misleading statement since the frequency averages can be calculated from the phase differences. The algorithm is exactly the same as integrating the frequency error is the same as calculating the phase change accumulated over the averaging time. Warrens implementation improves on the original NIST implementation by oversampling. Actually it degrades the simplicity and accuracy of the NIST implementation by replacing the integration inherent when using the counter and VFC with an approximation to the required frequency integral. Fortunately the accuracy can largely be recovered by using the appropriate signal processing algorithms. Think so? Lets see how well the VFC does at 1 ms and $10.00. I have NO trouble doing 1 sec integration, with errors that are far less than theirs. ( see end item) Yet again you seem to miss the point, cost is not relevant to the discussion of the correct signal processing technique. Why Warren omits this crucial step when all it requires is a little digital signal processing as all the required information is available from the sampled EFC voltage remains a mystery. Hay, The tester logs raw data, if you want to make a S/W filter to run the Raw data thru go for it, BUT before you waste your time, you should try and find at least one case that the current system does not work on. ( see end item) Yet again you demonstrate a profound lack of understanding of higher order numerical integration techniques. Its not that the method cant be easily fixed so that it produces accurate ADEV measures for an extremely wide range of sources with divergent phase noise spectra, its the extreme reluctance to do the signal processing correctly (its not that this even incurs extra hardware costs) that is perplexing. See above and Send me the easily fixed S/W. I'm ready to try it. The first test is to make sure it does not brake what is already working, then I can try it on anything that the existing S/W does not work on. Oh yet, that is going to be a bit of a problem, there is no known device the existing software does not work on. You'll need to send that along with the Software. How about hydrogen masers and cryogenic sapphire resonators? this is the END ITEM If you want to save a whole lot of time and not do extra S/W etc, I'll just make the oversample to tau0 ratio larger. That will fix any integration and phase noise and errors that you can come up with. If you do not understand why that is the case, then you still have no understanding of what I'm doing. Lets forget all the other BS for now, and just concentrate on the single statement. I'll increase the oversample to tau0 ratio, that will fix it That isnt always even possible or even a cost effective option. Another problem with the tight PLL method is that the PLL bandwidth is limited by the variable bandwidth of the EFC circuit (a few kHz for a 10811A). Thus accurate operation down to Tau =1ms may be somewhat problematic. If you think it is false, you really do need to go back to oversampling school. If you admit it is true, then we really do not have much else to talk about, because it fixes all your present concerns. It is now as simple as that. To discuss anything else is a total waste of time. Which is your way of saying that you don't understand the alternative more accurate methods and won't consider them. ws *** Bruce - Original Message - From: Bruce Griffiths bruce.griffi...@xtra.co.nz To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Thursday, June 03, 2010 12:27 AM Subject: Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A Steve Rooke wrote: On 3 June 2010 15:46, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: WarrenS wrote: As Bruce says It remains a mystery to him why this works. It doesnt, it only appears to in a very restricted set of circumstances. Bruce, I don't understand you, when presented with visual evidence that this method works you still deny it. What
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
WarrenS wrote: Bruce Posted Rectangular integration isn't particularly accurate or efficient, better techniques exist. True, but in this case it is the easiest and at these speeds, efficiency is not a big concern, It is made up for with faster oversampling. and it is obvious so far, better is not needed here, this is 'Good enough'. (and that answers your other question, why don't I do it better.) Why do that when its so easy to do much better? You do bring up an interesting point. There is lots of things that could be (and have been) done better than on that simple one IC BB circuit that was tested, and yet it was good enough to match the TSC 5120A pretty much point for point over the whole tau range and ADEV range.(limited only by it's Ref Osc.) Think KISS, enought said. You've entirely missed the point, such errors need to be quantified not swept under the carpet. Trapezoidal integration is almost as simple as rectangular integration and comes at low cost. However if you look at equation 44 in the paper I cited an even better technique is described. Try to desist from the pathetic attempts at insults I don't know what that even means, but sorry about the oversampling comments. It did seem you did not know what that was or at least its advantages when it comes to simplifying things. Ditto on the Phase and Freq differences comment, which I fear still may be the case. Sure if one has a sufficiently high oversampling factor crude approximations may work reasonably well. However one ought to strive to do better particularly when its easy to do so and requires no additional hardware. One doesn't always have the luxury of an extremely high oversampling ratio. A better filter than a single pole RC filter may also be required to take full advantage. You seem to be unaware of just how easy it is to create a dataset for which any given algorithm will fail catastrophically. True, I'm unaware of ANYTHING Normal that will make this fail. send me something, It'll be fun to try it, If you can break it, I can fix it. Lets start small, give me any two numbers, I'll give you the average Now three, then four, how far do you want to go? I'm sure I can still give you the average value. Now for the big test, can I give you the average of several of the previous averages, This is not going to be a problem either. I can give you the average for any reasonable numbers that could be presented to the ADC in normal operation thru the restricted PLL BW and the BW filter. That IS about ALL there is to it. It needs to give the AVERAGE Frequency of the oversampled average frequencies. Then it needs to give the average of these averages. As long all samples are taken at the same rate, it works good. (Average = Sum_nSamples / n ) The oversampled has to be done fast enough and with the appropriate B/W limit so there is no dead time or aliasing or significant change happens during one sample period. The other H/W takes care of that. Oversampling TC filter. To understand why that is all that is needed, one need to only look at what basic Allan deviation is. Allan deviation is the Average of the neighboring frequency differences that have been averaged for a given length of time. That length of time is called tau. Actually you should be using fractional frequency differences. That is needed then is to find (the Average_Freq over a tau period) - (the Average_Freq over the next same length time period). Do some squaring of the differences and some more averaging and scaling and some sq root and out pops an ADEV answer. The important point is that it ALL just starts with the Average Frequency over a period of time called tau. If you can get accurate average freq over tau0 time then any standard Allan calculation S/W can turn it into ADEV and at any tau. ws *** Bruce WarrenS wrote: Bruce posted The RC filter doesn't accurately integrate the frequency difference over time interval Tau0. For you to even state that means you still have NO idea what I'm doing, It is getting sort of sad. Correct the RC filter is not an integrator, it is used for the combination Bandwidth and anti-aliasing filter. It is the oversampling average that does the integration. How? Rectangular integration isnt particularly accurate or efficient, better techniques exist. What would explain a lot is, if you do not know what oversampling even is? Try to desist from the pathetic attempts at insults as they merely distract from the real questions about the signal processing techniques adopted. You need to get yourself a refresher course on the advantages of oversampling to do integration, brick wall filtering, anti-aliasing and why a single RC works just fine for integration when oversampling is used and why you don't need anything but simple averaging of sum n samples /n when oversampling is
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
, The tester logs raw data, if you want to make a S/W filter to run the Raw data thru go for it, BUT before you waste your time, you should try and find at least one case that the current system does not work on. ( see end item) Yet again you demonstrate a profound lack of understanding of higher order numerical integration techniques. Its not that the method cant be easily fixed so that it produces accurate ADEV measures for an extremely wide range of sources with divergent phase noise spectra, its the extreme reluctance to do the signal processing correctly (its not that this even incurs extra hardware costs) that is perplexing. See above and Send me the easily fixed S/W. I'm ready to try it. The first test is to make sure it does not brake what is already working, then I can try it on anything that the existing S/W does not work on. Oh yet, that is going to be a bit of a problem, there is no known device the existing software does not work on. You'll need to send that along with the Software. How about hydrogen masers and cryogenic sapphire resonators? this is the END ITEM If you want to save a whole lot of time and not do extra S/W etc, I'll just make the oversample to tau0 ratio larger. That will fix any integration and phase noise and errors that you can come up with. If you do not understand why that is the case, then you still have no understanding of what I'm doing. Lets forget all the other BS for now, and just concentrate on the single statement. I'll increase the oversample to tau0 ratio, that will fix it That isnt always even possible or even a cost effective option. Another problem with the tight PLL method is that the PLL bandwidth is limited by the variable bandwidth of the EFC circuit (a few kHz for a 10811A). Thus accurate operation down to Tau =1ms may be somewhat problematic. If you think it is false, you really do need to go back to oversampling school. If you admit it is true, then we really do not have much else to talk about, because it fixes all your present concerns. It is now as simple as that. To discuss anything else is a total waste of time. Which is your way of saying that you don't understand the alternative more accurate methods and won't consider them. ws *** Bruce - Original Message - From: Bruce Griffiths bruce.griffiths at xtra.co.nz To: Discussion of precise time and frequency measurement time-nuts at febo.com Sent: Thursday, June 03, 2010 12:27 AM Subject: Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A Steve Rooke wrote: On 3 June 2010 15:46, Bruce Griffithsbruce.griffiths at xtra.co.nz wrote: WarrenS wrote: As Bruce says It remains a mystery to him why this works. It doesnt, it only appears to in a very restricted set of circumstances. Bruce, I don't understand you, when presented with visual evidence that this method works you still deny it. What visual evidence?? There is no proof that the technique works well in every case. Only that for the range of Tau tested and for the particular source pair used that it appears to. Not one of my best skills, but I'll try to explain it once again. Now that they see it works, maybe someone else will be able to put this into words that Bruce will be able to finally understand. The only requirement needed for the Frequency data log to be give correct ADEV readings, is to get good, Averaged, integrated, Frequency data, with no dead time, and no aliasing, over the tau0 time period. Each Tau0 Frequency sample is ideally completely independent from all others. If it can do one right then it can get them ALL right. In a single tau0 sample there is NO SUCH THING as a certain type of long term noise, Just the average freq over that single time period. Misleading as usual, your knowledge of statistics is woefully inadequate leading to incorrect conclusions. Well, what are are the woefully inadequate conclusions then? Please give us your full reasoning. A simple example is that for a small number of samples a stability metric like the ordinary (unfiltered) phase variance standard deviation may appear to be stable, whereas with a sufficiently large number of samples the instability of the metric itself becomes evident whenever divergent noise processes like flicker phase noise, random walk frequency noise are present. /Each Tau0 Frequency sample is ideally completely independent from all others. / The above statement is incorrect as the finite bandwidth necessarily imparts a correlation between samples they can only be strictly independent if the bandwidth is infinite. /In a single tau0 sample there is NO SUCH THING as a certain type of long term noise, Just the average freq over that single time period. / The above statement imparts no useful information. It would be much easier and less bandwidth wasted if the circuit schematics and useful documentation on the algorithms employed were available. Extracting any useful information seems
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
Steve Rooke wrote: Bruce, On 3 June 2010 19:27, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Bruce, I don't understand you, when presented with visual evidence that this method works you still deny it. What visual evidence?? There is no proof that the technique works well in every case. Only that for the range of Tau tested and for the particular source pair used that it appears to. I have already commented on this in another thread but to reiterate. The test that John performed that for a range of Tau that was possible to be calculated for the given measurement period, both methods produced the same results for each and every value of Tau, not for a single value of Tau. Well, what are are the woefully inadequate conclusions then? Please give us your full reasoning. A simple example is that for a small number of samples a stability metric like the ordinary (unfiltered) phase variance standard deviation may appear to be stable, whereas with a sufficiently large number of samples the instability of the metric itself becomes evident whenever divergent noise processes like flicker phase noise, random walk frequency noise are present. OK, we need to run the test for a longer period but, as John has indicated, he is not able to devote any further time to this. That fact does not mean that this method has no value, it is just like every other area of theoretical physics, we can never prove it true so we try to look at ways of proving it false. The important words there are proving it false. Lots of stuff deleted. Lets explore frequency measurement in a way that we all can understand. No oscillator can be measured in isolation, it has to be measured against another standard oscillator. Conventional frequency measurement is performed by counting the number of cycles of the unknown oscillator over a known period or gate time. This averages the unknown frequency over the gate time so that an instrument built using this method can only provide an accuracy down to a single cycle over the number of cycles that are counted during the gate time. What this means is that the conventional method of frequency measurement averages the measured value and is subject to an error of the period of the frequency. So how does Warren's system measure the frequency. Using the tight-PLL method the feedback voltage controlling the reference oscillator is constantly tracking any difference between both oscillators. If measurements of the PLL feedback are only taken at the required Tau frequency, the result would only show the instantaneous value which would be equivalent to basically measuring the period of the last (or last few, depending upon the Tc of the PLL feedback filter) cycles. By oversampling, IE. taking many measurements during the desired period of minimum Tau, these measurements can be averaged to produce the averaged frequency of that Tau period. This means that the PLL filter loop can be made much faster to keep the unknown and reference oscillators tightly tracking. Without some dampening in the PLL loop, the circuit would become unstable and quite unusable. All that is required in this filter is a very simple integrator as it's Tc is faster than the oversampling rate which is in turn faster than the Tau period. Only in your imagination. One would assume that this method only works when Warren does it as his imagination is required for it to work, but wait, John Miles has managed to get point for point identical data against a TSC, how can that be Bruce? Please give answers, not insults. Read the following paper: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf This is a fine paper but is only relevant to the method they are using for AVAR determination. As I have explained above, oversampling enables the implementation of a very simple filter in the PLL loop. which shows the relationship between AVAR etc, filters and the ordinary phase variance. The paper also outlines the techniques that should be used with the sampled frequency difference data from a tight PLL. Agreed, but they had not thought about oversampling.. Again, it it does not work, how come the evidence shows that it does, how do you explain that Bruce? The evidence doesn't show this at all. It merely indicates that for the devices tested that the phase noise spectral components in the region where the filter responses of the ADEV and WDEV differ (its not ADEV so it shouldnt be labelled as such) dont appear to be significant for the 2 sources compared and the tau range over which the testing was done. OK, for all intents and purposes ADEV measurements are only carried out on a limited variety of oscillators as it would be pointless to, say, perform such a measurement on an LC disciplined oscillator. So we are looking at xtal, Rb, Cs and Hm at the moment. For practical purposes the xtal has good
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
John Miles wrote: For those following this strange and wonderful saga: http://www.ke5fx.com/tpll.htm -- john, KE5FX ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. The problem with that page is that you show the original NIST implementation which actually produces valid ADEV measures whereas Warren's implementation omits the crucial integration/averaging (his figurative handwaving antics don't change this) and hence actually has a different phase noise frequency response than that of the filter implied by the definition of AVAR. Why Warren omits this crucial step when all it requires is a little digital signal processing as all the required information is available from the sampled EFC voltage remains a mystery. The method as implemented by Warren produces a frequency stability metric which may be useful for comparing the stability of some sources, however it does not measure ADEV. Under a restricted set of circumstances such as when white phase noise or drift dominate the measures so calculated my be close to the measured ADEV obtained by a method wth the correct response to the various phase noise frequency components, however this doesnt mean that the measures are actually ADEV measures it merely means that the phase noise frequency components in the region where the frequency response of the 2 methods differ significantly, are not significant. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
WarrenS wrote: As Bruce says It remains a mystery to him why this works. It doesnt, it only appears to in a very restricted set of circumstances. Not one of my best skills, but I'll try to explain it once again. Now that they see it works, maybe someone else will be able to put this into words that Bruce will be able to finally understand. The only requirement needed for the Frequency data log to be give correct ADEV readings, is to get good, Averaged, integrated, Frequency data, with no dead time, and no aliasing, over the tau0 time period. Each Tau0 Frequency sample is ideally completely independent from all others. If it can do one right then it can get them ALL right. In a single tau0 sample there is NO SUCH THING as a certain type of long term noise, Just the average freq over that single time period. Misleading as usual, your knowledge of statistics is woefully inadequate leading to incorrect conclusions. The crucial integration/averaging to get good tau0 data, that Bruce can not see for some unknown reason, is done Only in your imagination. with an analog filter set to about the Tau0 Freq and by oversampling at about about a 10 to one ratio, and averaging the oversampled frequency readings down to tau0. That doesn't work as it has the wrong transfer function. It is not perfect, but plenty close enough for the plot to match the output of the TSC 5120A over the whole tau range. There are a few other subtle details on how to insure that aliasing and over filtering do not become a problem, but first things first, one needs to understand how the integration is being done. Sloppy and misleading explanation as usual. The integration secret (which is no secret to anyone but Bruce) is to analog filter, Oversample, then average the Frequency data at a rate much faster than the tau0 data rate. Which again is misleading as you specify neither the averaging method nor the analog filter. That alone should be enough information for any knowledgeable designer to understand. Its not and you should know that it isnt. You draw conclusions that are neither supported by measurement nor theory. ws ps) Do note, I'm working with Frequency here and not phase, that may be what is confusing some. When will you understand that phase differences and differences of average frequency (unit weight to frequency measures over the sampling interval zero weight outside) are equivalent. *** The problem with that page is that you show the original NIST implementation which actually produces valid ADEV measures whereas Warren's implementation omits the crucial integration/averaging (his figurative handwaving antics don't change this) and hence actually has a different phase noise frequency response than that of the filter implied by the definition of AVAR. Why Warren omits this crucial step when all it requires is a little digital signal processing as all the required information is available from the sampled EFC voltage remains a mystery. The method as implemented by Warren produces a frequency stability metric which may be useful for comparing the stability of some sources, however it does not measure ADEV. Under a restricted set of circumstances such as when white phase noise or drift dominate the measures so calculated my be close to the measured ADEV obtained by a method wth the correct response to the various phase noise frequency components, however this doesnt mean that the measures are actually ADEV measures it merely means that the phase noise frequency components in the region where the frequency response of the 2 methods differ significantly, are not significant. Bruce * John Miles wrote: For those following this strange and wonderful saga: http://www.ke5fx.com/tpll.htm -- john, KE5FX ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Notes on tight-PLL performance versus TSC 5120A
John Miles wrote: The integration secret (which is no secret to anyone but Bruce) is to analog filter, Oversample, then average the Frequency data at a rate much faster than the tau0 data rate. Which again is misleading as you specify neither the averaging method nor the analog filter. I can't speak for the analog side as I never saw a schematic of the PLL, but it may be worthwhile to point out that the averaging code in question is in SOURCE_DI154_proc() in ti.cpp, which is installed with http://www.ke5fx.com/gpib/setup.exe . This is my code, not Warren's. It does a simple boxcar average on phase-difference data, the same as my TSC 5120 acquisition routine does. Previous tests indicated that simple averaging yields a good match to most ADEV graphs on TSC's LCD display, so I used it for the PLL DAQ code as well. I also tried a Kaiser-synthesized FIR kernel for decimating the incoming TSC data, but found that its conformance against the TSC's display was worse than what I saw with the simple average. More work needs to be done here. When will you understand that phase differences and differences of average frequency (unit weight to frequency measures over the sampling interval zero weight outside) are equivalent. One subtlety is the question of whether to average (or otherwise filter) the DAQ voltage readings immediately after they're acquired and linearly scaled to frequency-difference values, versus after conversion of the frequency-difference values to phase differences. I found that the best agreement with the TSC plots was obtained by doing the latter: val = (read and scale the DAQ voltage) // val is now a frequency difference // averaging val here yields somewhat higher // sigma(tau) values in the first few bins // after tau0 val = last_phase + (val / DI154_RATE_HZ); last_phase = val; This appears to use a rectangular approximation to the required integral. A trapezoidal or even Simpson's rule integration technique should be more accurate for a given sample rate. One could even try a higher order polynomial fit to the sample points, however this isnt the optimum technique to use. If one uses WKS interpolation to reconstruct the continuous frequency vs time function and integrates the result for a finite time interval (Tau0) then one ends up with a digital filter with infinite number of terms. Since an infinite number of samples is required to do this using a suitable window function is probably advisable. The paper (below) illustrates how AVAR etc can be calculated from the sampled frequency difference data using DFT techniques: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf // val is now a phase difference // averaging val here matches the TSC better The difference is not huge but it's readily noticeable. This is subtly disturbing because the RC filter before the DAQ *does* integrate the frequency-difference data directly. If it's correct to band-limit the frequency-to-voltage data in the last analog stage of the pipeline, it should be correct to do it in the first digital stage, I'd think. The RC filter doesnt accurately integrate the frequency difference over time interval Tau0. Further complicating matters is the question of whether the TSC 5120A's filtering process is really all that 'correct,' itself. When they downsample their own data by a large fraction, e.g. when you select tau0=100 msec / NEQ BW = 5 Hz, there is often a slight droop near tau0 that does not correspond to anything visible at higher rates. To some extent we may be attempting to match someone else's bug. This is the result of the choice of the low pass filter bandwidth made by the designers. The filter bandwidth increases as Tau0 decreases. The traditional analyses of the dependence of AVAR on bandwidth of this filter assume a brickwall filter. At any rate I've run out of time/inclination to pursue it, at least for now. The SOURCE_DI154_proc() routine in TI.CPP is open for inspection and modification by any interested parties, lines 6753-7045 in the current build. :) Warren has his hardware back now, and would presumably be able to try any modifications. -- john, KE5FX ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Digital tight PLL method
Steve Rooke wrote: On 28 May 2010 07:42, Bruce Griffithsbruce.griffi...@xtra.co.nz wrote: Steve Rooke wrote: On 28 May 2010 04:40, David Martindaledave.martind...@gmail.comwrote: Hmm. From here in Vancouver Canada, the name resolves to the same address, pings fail, and the given URL gets me the web page. Try using the top-level page address: http://tf.nist.gov/phase/Properties/main.htm (Looks like the whole set of pages is worth reading anyway). I've just discovered that if I use Tor it works fine. Perhaps it does not like New Zealand :( Steve Dave Only those in the South Island. We call it The Mainland here mate! Steve Bruce I debated using that term but others on the list wouldnt be aware of it. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Digital tight PLL method
Ulrich Close in spurs generated by the synthesiser may also be problematic. One feature being that the spur levels will depend (in a complex way) on the synthesiser output frequency. One of the first problems to solve is making the close in spurs sufficiently low. Another problem is to ensure that the synthesiser output is phase continuous (not a problem with DDS but close in spurs may be). Bruce Ulrich Bangert wrote: Warren, you are not the only person to have ideas like this! I managed to get me a Stanford Research DS345 generator which gives 1E-6 Hz frequency resolution for any frequency below 30 MHz (Can be locked to any 10 MHz reference). At 10 MHz this resembles a relative resolution of 1E-13. I used this generator in a digital pll where the phase error was measured by a DBM and a a HP3457. The digital PLL was a simple script written with my EZGPIB utility which controlled the DS345 and read the HP3457 via IEEE488. The main difference to your analogue solution is that it delivers a frequency measurement value immediately (= the current setting of the DS345) without any knowledge needed about the mixer's phase gain properties. And it is not limited to a certain frequency. Of course, the generator may be exchanged by an DIY DDS and the multimeter may be exchanged against a DIY A/D converter. Injection locking is not a topic with the DDS circuit. Nevertheless my measurement were not exactly encouraging. May be that I missed to apply the important math that Bruce has been suggesting in the discussion with you. All the stuff is on my workbench and is ready to use. May be I give it another try. Best regards Ulrich Bangert -Ursprungliche Nachricht- Von: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] Im Auftrag von WarrenS Gesendet: Montag, 24. Mai 2010 18:49 An: John Miles; Tom Van Baak; Discussion of precise time and frequency measurement Betreff: [time-nuts] Digital tight PLL method Concerning the simple, $10, Low cost, Tight PLL method of doing ADEV. If you accept that the measurement is going to be limited by the Reference Osc, Then for Low COST and SIMPLE, with the ability to measure ADEVs at very low levels, Can't beat a simple analog version of NIST's Tight Phase-Lock Loop Method of measuring Freq stability. http://tf.nist.gov/phase/Properties/one.htm#oneoneFig 1.7 Here is some more information on the subject that may help inspire some of the great minds out there. In spite of all the unjustified criticism, the latest test will show, at least to the more open minded nuts, There is NOTHING inherently wrong with the tight PLL method as I have done it. It gives about as good of answers as anything out there. As I've implemented it, there are some disadvantages, because there is just so much one can do with a single Op amp design. If one does the calculation they will also see the OP amp is not a limiting factor in the performance of this method. AS I have said before, the disadvantage of my simple BB version that was tested, is that it is limited by the Ref Osc and the way it's freq is modified. The accuracy is limited by the fact the first simple BB version I built is an all analog system. That is solely because the frequency control I used on the simple version is the analog EFC input of the reference Osc. I've also pointed out, that is not a limitation of the method, there are solutions for that. Now I'm amazed that no one has had a New inspiration. Maybe a more direct approach will help some to see the next logical step. Using the same basic tight PLL method, make some of the unit digital. Do not modify the freq of the reference osc with analog, GET it yet? That way the device would be half digital without any of the analog shortcoming or the need to physically change the reference freq. Do I really need to explain more? Have fun ws *** ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Z3805 utility, Was: AW: (no subject)
Didier Juges wrote: Bill, I think you got it backwards. +/- 12V is typical for RS-232, 0/+5V is for RS-422 and RS-485. No RS-232 receiver should be damaged with +/- 12V or even +/- 15V because that is their normal operating voltage. Also, RS-422 and RS-485 have something like 25V common mode tolerance (not sure what the actual spec is there,) so that the RS-422 and RS-485 receivers should not be damaged by 15V either. The RS-422 drivers are pretty low impedance, while the RS-232 drivers are current limited, so I don't think that connecting an RS-232 driver into an RS-422 driver will damage either. Picking one RS485 receiver (ADM1485) at random the receiver absolute maximum (no damage) input range is -14V to +14V. The RS485 receiver operating common mode range is -7V to +12V. RS422 receivers have an input operating range of -7V to +7V. The no damage RS422 receiver input ratings may be higher. However, most recent (10 years?) RS-232 receivers will work with a 0/+3V or 0/+5V input, conveniently having a threshold a few 10's or 100's of mV above ground, even though the original RS-232 spec required receivers that work with as low as +/- 3V, and drivers that deliver +/- 9V minimum. Many commercial systems use +/- 5V drivers for RS-232 (BB Electronics sells a lot of converters with these voltages). This is a deliciously sloppy spec that nobody has met in the last 25 years probably, yet works most of the time. The one thing to avoid is to short an RS-422 (or RS-485) driver to ground, as that can actually cause damage, maybe not every time, but definitely not recommended. These have relatively high current output capability to drive long lines. Didier KO4BB Bruce -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Bill Hawkins Sent: Saturday, May 22, 2010 3:09 PM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] Z3805 utility, Was: AW: (no subject) When all else fails, get out the voltmeter. Do you have power to the antenna? Is it the right voltage? All the way to the antenna? What volts are on pins 2 or 3 relative to pin 7 in the comm connector? If you see 12 volts, that's RS-422. You may have burned out your computer's serial port. If you see less than 5 volts, that's RS-232 and all should be well, unless you see zero volts. I may have the RS-xxx volts somewhat off because my memory isn't what it used to be. The guy you bought it from should be able to help with comm basics. Bill Hawkins -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert Benward Sent: Saturday, May 22, 2010 2:08 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Z3805 utility, Was: AW: (no subject) Hi All, I hooked everything up and I still get nothing. I can't seem to establish communications with the Z3805. I tried a null modem as well, in case the cable (supplied) was wired with the wrong connector gender. I see a green blinking light inside, it he left rear corner of the box. Everything is warm, but nothing else. Any ideas? Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Dual Mixer
WarrenS wrote: Bruce Good, It does seem like we are finally making some good progress. You now seem to acknowledge that my tester could work if I integrate. You now seem to acknowledge that I am integrating by using a filter. In a sampled data system integration is equivalent to a filter but not just any arbitrary low pass filter. The errors in your method are explicitly spelled out in the paper I gave the link to: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf In this paper xi is a phase sample and yi is a frequency sample. I acknowledge that my integration method is not perfect, BUT it is simple and good enough. Not yet proven nor quantified. It would seem the only issue left is to show you just how good of answers my integration method gives. At least now we are JUST talking about what the S/W needs to do. Hopefully you now see that the hardware is adequate. What would you consider an acceptable error band, 3 dB, 1 dB, 0.1 dB? Pick a number zero. The answer depends on how long one is willing to spend making the measurements. Certainly 0.1dB or better would require heroic efforts to demonstrate. Since the error will also depend on the phase noise spectra of the oscillators being compared a single figure answer isnt feasible. However for the case where white phase noise dominates the error should be not more than 1dB but potentially much less. The errors due to digital signal processing should be at least an order of magnitude lower. For a typical high speed data log taken at say 1 K samples per second, one would generally run a quick test with maybe a minute's worth of data. That would provide enough data to give a good tau plot up to about 10 seconds. That's a rather sweeping statement given that no estimates of the contribution to measurement noise due to the finite number of samples has been made. The maximum usable tau for a given record length depends on the maximum acceptable error due to the finite number of samples. Now if you can supply me with a 60K data log with any type of reasonably typical noise that you want to include in it I'll show you how close my approximate Integration comes to your perfect integration. You can't because your method of perfect integration isnt and its errors cannot be made sufficiently small with so few samples. I can set this up to do as many times as you want, until I have demonstrated by example that it is close enough, for every data log case that you will provide. Near enough IS good enough for me and most Nuts. Quantify near enough else all is just noise. As John pointed out, this is measuring noise. One is not going to get the exact same answer twice in a row anyway. My answer will not be perfect, but it will be simple and fast and easy and below the noise uncertainty band. Your turn to put a data log where your math is. Do try and remember I'm working with Frequency and not phase. Thats idle speculation as you havent quantified anything at all. The repeatability of the measurements needs to be quantified. BTW. just a heads-up warning to be fair. I have set up this situation so that I can not loose. Its actually almost trivial to produce a set of samples for which any given method will fail. Doing so is an unproductive exercise. If you want to setup your own situation go for it. I'll see if I can do it. Only requirement is that it should be broken down into no more than 60K sample sizes max for each test at the start. After I pass that, if you want to go for millions of samples or whatever, fine as long as I can read the text data log file. ws Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Dual Mixer
As long as one is aware that your method (as implemented by you) doesn't actually measure Allan variance, it may be useful for comparing the relative stability some sources for small Tau (unfortunately the range of Tau for which the method may produce useful results depends on the phase noise characteristics of the sources being compared). To measure AVAR the technique has to have the same response to all phase noise spectral components as does AVAR. Since you do not integrate/average the frequency measures the phase noise response of the method is not identical to that used in calculating AVAR. This technique probably works best when white phase noise dominates the phase noise spectral region of interest (usually for small Tau). For those who can follow the theory, the following paper shows how the above method is affected by aliasing etc: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf The paper also shows how the required integration (needed to actually measure AVAR) can be approximated from the discrete sample sequence. Alternatively one could avoid the numerical integration by replacing the ADC with a zero deadtime (ie not a dual slope converter. A multislope algorithm like that used in the 34401A (but not the 3458A) should work as the signal is integrated continuously) integrating ADC. One possibility is to use a VFC as NIST did when they used this technique some decades ago. Of course, the classical DMTD setup undersamples the phase noise spectrum and thus may suffer from aliasing artifacts. Such aliasing artifacts have no significant effect when the phase noise spectrum is flat. Bruce WarrenS wrote: For the Really cheap time nuts, It sounds like Bert Kehren has done a great Job building a Dual Mixer tester. There are other simpler, less standard ways to get good data for Allan Variance and small frequency differences. My VERY simple $10.00 analog tight PLL Tester BB (Previoulsy posted) pretty much accomplishes the same goals as his, and it can do 1e-13 in a second, and 1e-11 in 10ms (limited of course by the single reference Oscillator used) A simple test that most can do at home, and still challenges the best high end testers out there is Tom's the swinging Oscillator test. http://www.leapsecond.com/pages/10811-g/ (The results from my PLL tester is attached) ws ** - Original Message - From: ewkeh...@aol.com To: time-nuts@febo.com Sent: Tuesday, May 11, 2010 7:02 AM Subject: [time-nuts] Dual Mixer The Dual Mixer project is nearing completion. Let me refresh every ones memory as to my goals. a) Total cost less than $ 200 b) 1 E-13 with a one second offset c) use parts attainable by every one d) easy to assemble only a few surface mount parts e) a five channel counter that yields 1 E 15 resolution and interfaces directly to a PC via RS232 or USB f) A counter that also gives you instant frequency difference at the sample rate, not only Allan Variance g) Modular so one can use only the Dual Mixer h) Modular so one can use multiple units to do simultaneous comparison of more than two oscillators. i) Isolation between D/M and counter so that the counter can be powered by the PC USB port I am happy to report that all goals have been accomplished, attached is a picture of the D/M, limitation of the file size does not allow me to attach an actual board picture, but if you contact me direct I will send you one, the final board is actually nicer since the first layout had to accommodate several variances. The D/M part leans heavy on the original NIST unit with a few substitutions and recommendations from Bob Camp. Also beside Opto Couplers SN65LVDS1's have been included for those that want to use other counting methods. Selection of filter capacitors allow the use at other offset frequencies such as 10 and 100 Hz. The D/M fits in a standard 74 X 111 X 20 mm Euro case and the counter can be stacked below or next to it using the Opto Isolators as the inter connect. The SYPD-1's fit right on the board but connections are included to use the HP 10514 A. As a matter of fact removing the HP mixer board from its housing fits it nicely on the board and every thing is still inside the housing. The counter will handle 1 an 10 Hz offset with a 1 E 14 resolution at 10 Hz. Thanks to Richard Mc Corkle we have great drawings and code, available to every one. Code, drawings, list of material and PC board layouts and its file, will be available to every one once the project is completed. I need help in the following areas a) help me create a nice set of drawings that are computer generated something I am not able to do b) create the computer program that takes the output of the counter board and allows Allan Variance plots, frequency difference and dual temperature readings and plots using RS232 and USB. c) an independent test by a third party. As I said previously, I am not getting
Re: [time-nuts] Dual Mixer
Warren So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Bruce WarrenS wrote: Bruce Before we go around again and discuses what my simple tester can and can not do and why, It would be helpful if you would take the time to better understand how it works and why it works the way I have done it. You really should try one yourself if you can't see why it works. You are going to be surprised and embarrassed at how good it works. Why you're at it, try the swing test with anything you have. Let me know how that goes. I'm not saying that may tester will match someone's Latest ever changing NEW idea of what the correct AVAR should be, After all it just Logs correct, integrated, Freq difference data of ANY noise type and does it without adding any dead time or aliasing all by uisng pretty much using ANY ADC capability of over sampling at the tau Zero rate. If one then uses the data log with something like the classic Stable 32 S/W or Ulrich's Plotter, it gives is the exact same results as other methods costing much much more, over the whole tau range. This is limited only be its reference oscillator (Same way that all others are limited of course, Doen't get much better than that). If that is not good enough for you, them you need to discuss the results with Symmetricon and others that give the same answer as mine, not me. If for some reason you want to set one up wrong so that it matches the results of some other special instrument, I'd be glad to tell you how to have it add back in the dead time or aliasing artifact problems or whatever else you would like it to do wrong, that it presently does correctly. ws ** Bruce wrote As long as one is aware that your method (as implemented by you) doesn't actually measure Allan variance, it may be useful for comparing the relative stability some sources for small Tau (unfortunately the range of Tau for which the method may produce useful results depends on the phase noise characteristics of the sources being compared). To measure AVAR the technique has to have the same response to all phase noise spectral components as does AVAR. Since you do not integrate/average the frequency measures the phase noise response of the method is not identical to that used in calculating AVAR. This technique probably works best when white phase noise dominates the phase noise spectral region of interest (usually for small Tau). For those who can follow the theory, the following paper shows how the above method is affected by aliasing etc: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf The paper also shows how the required integration (needed to actually measure AVAR) can be approximated from the discrete sample sequence. Alternatively one could avoid the numerical integration by replacing the ADC with a zero deadtime (ie not a dual slope converter. A multislope algorithm like that used in the 34401A (but not the 3458A) should work as the signal is integrated continuously) integrating ADC. One possibility is to use a VFC as NIST did when they used this technique some decades ago. Of course, the classical DMTD setup undersamples the phase noise spectrum and thus may suffer from aliasing artifacts. Such aliasing artifacts have no significant effect when the phase noise spectrum is flat. Bruce * WarrenS wrote: For the Really cheap time nuts, It sounds like Bert Kehren has done a great Job building a Dual Mixer tester. There are other simpler, less standard ways to get good data for Allan Variance and small frequency differences. My VERY simple $10.00 analog tight PLL Tester BB (Previously posted) pretty much accomplishes the same goals as his, and it can do 1e-13 in a second, and 1e-11 in 10ms (limited of course by the single reference Oscillator used) A simple test that most can do at home, and still challenges the best high end testers out there is Tom's the swinging Oscillator test. http://www.leapsecond.com/pages/10811-g/ (The results from my PLL tester is attached) ws ** - Original Message - From: EWKehren at aol.com To: time-nuts at febo.com Sent: Tuesday, May 11, 2010 7:02 AM Subject: [time-nuts] Dual Mixer The Dual Mixer project is nearing completion. Let me refresh every ones memory as to my goals. a) Total cost less than $ 200 b) 1 E-13 with a one second offset c) use parts attainable by every one d) easy to assemble only a few surface mount parts e) a five channel counter that yields 1 E 15 resolution and interfaces directly to a PC via RS232 or USB f) A counter that also gives you instant frequency difference at the sample rate, not only Allan Variance g) Modular so one can use only the Dual Mixer h) Modular so one can use multiple units to do simultaneous comparison of more than two oscillators. i) Isolation between D/M and counter so that the counter can be
Re: [time-nuts] Dual Mixer
You cannot approximate the sinc function frequency response of an ideal integrator with an arbitrary low pass filter. Your scheme will tend to misbehave (in that it will produce anomalous ADEV estimates) when flicker phase noise is significant. You actually need to use an analog low pass filter (or its equivalent) and an integrator to produce useful ADEV measures Bruce WarrenS wrote: (My apologies to all, this is a game Bruce and I play every time I bring up my simple tester.) Bruce wrote: So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Yes, I integrate/average just the same as I have always done it from day one. Did you finally understand how the integration works using most any ADC? Hint: it's done with oversampling the tau zero time. (and a LP filter set to a value above the tau zero but below the oversamping rate) The VERY SAME thing I have been trying to tell you from day one, something that you have chosen to ignore. The very original Block diagram that I posted shows it, if you need more information. ws *** Warren So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Bruce * WarrenS wrote: Bruce Before we go around again and discuses what my simple tester can and can not do and why, It would be helpful if you would take the time to better understand how it works and why it works the way I have done it. You really should try one yourself if you can't see why it works. You are going to be surprised and embarrassed at how good it works. Why you're at it, try the swing test with anything you have. Let me know how that goes. I'm not saying that may tester will match someone's Latest ever changing NEW idea of what the correct AVAR should be, After all it just Logs correct, integrated, Freq difference data of ANY noise type and does it without adding any dead time or aliasing all by using pretty much using ANY ADC capability of over sampling at the tau Zero rate. If one then uses the data log with something like the classic Stable 32 S/W or Ulrich's Plotter, it gives is the exact same results as other methods costing much much more, over the whole tau range. This is limited only be its reference oscillator (Same way that all others are limited of course, Doesn't get much better than that). If that is not good enough for you, them you need to discuss the results with Symmetricon and others that give the same answer as mine, not me. If for some reason you want to set one up wrong so that it matches the results of some other special instrument, I'd be glad to tell you how to have it add back in the dead time or aliasing artifact problems or whatever else you would like it to do wrong, that it presently does correctly. ws ** Bruce wrote As long as one is aware that your method (as implemented by you) doesn't actually measure Allan variance, it may be useful for comparing the relative stability some sources for small Tau (unfortunately the range of Tau for which the method may produce useful results depends on the phase noise characteristics of the sources being compared). To measure AVAR the technique has to have the same response to all phase noise spectral components as does AVAR. Since you do not integrate/average the frequency measures the phase noise response of the method is not identical to that used in calculating AVAR. This technique probably works best when white phase noise dominates the phase noise spectral region of interest (usually for small Tau). For those who can follow the theory, the following paper shows how the above method is affected by aliasing etc: http://hal.archives-ouvertes.fr/docs/00/37/63/05/PDF/alaa_p1_v4a.pdf The paper also shows how the required integration (needed to actually measure AVAR) can be approximated from the discrete sample sequence. Alternatively one could avoid the numerical integration by replacing the ADC with a zero deadtime (ie not a dual slope converter. A multislope algorithm like that used in the 34401A (but not the 3458A) should work as the signal is integrated continuously) integrating ADC. One possibility is to use a VFC as NIST did when they used this technique some decades ago. Of course, the classical DMTD setup undersamples the phase noise spectrum and thus may suffer from aliasing artifacts. Such aliasing artifacts have no significant effect when the phase noise spectrum is flat. Bruce * WarrenS wrote: For the Really cheap time nuts, It sounds like Bert Kehren has done a great Job building a Dual Mixer tester. There are other simpler, less standard ways to get good data for Allan Variance and small frequency differences. My VERY simple $10.00 analog tight PLL Tester BB (Previously posted) pretty much accomplishes the same goals as his, and it can do 1e-13 in a second, and 1e-11 in
Re: [time-nuts] Dual Mixer
The results have so far only been shown to be useful when white phase noise dominates. When the phase noise is white almost anything can be made to produce a result that differs from ADEV by at scale factor. In practice its sometimes difficult to know over what range of Tau that the phase noise is in fact white. The various tests and comparisons that have been made or are underway are necessary but not sufficient proof of the usefulness of this technique. The phase noise frequency response of the technique is also required so that its limitations can be delineated. 1000 samples of a divergent noise process are insufficient, spreadsheet analysis of the millions of samples that are probably necessary is impossible/impractical. Using something like Matlab is probably necessary to achieve meaningful results. Bruce WarrenS wrote: OK, So, It is not perfect, but its simple and does give answers that are GOOD enough. At least you now understand if the integrator works then the tester works. So that we do not go down hill again after all this progress, If you would like to send me a data file of say 1000 + samples of any noise type of your choice I'll send you back an excel spread sheet to show the insignificant error that this integration method produces. ws You cannot approximate the sinc function frequency response of an ideal integrator with an arbitrary low pass filter. Your scheme will tend to misbehave (in that it will produce anomalous ADEV estimates) when flicker phase noise is significant. You actually need to use an analog low pass filter (or its equivalent) and an integrator to produce useful ADEV measures Bruce *** WarrenS wrote: (My apologies to all, this is a game Bruce and I play every time I bring up my simple tester.) Bruce wrote: So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Yes, I integrate/average just the same as I have always done it from day one. Did you finally understand how the integration works using most any ADC? Hint: it's done with oversampling the tau zero time. (and a LP filter set to a value above the tau zero but below the oversamping rate) The VERY SAME thing I have been trying to tell you from day one, something that you have chosen to ignore. The very original Block diagram that I posted shows it, if you need more information. ws *** Warren So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Bruce * WarrenS wrote: Bruce Before we go around again and discuses what my simple tester can and can not do and why, It would be helpful if you would take the time to better understand how it works and why it works the way I have done it. You really should try one yourself if you can't see why it works. You are going to be surprised and embarrassed at how good it works. Why you're at it, try the swing test with anything you have. Let me know how that goes. I'm not saying that may tester will match someone's Latest ever changing NEW idea of what the correct AVAR should be, After all it just Logs correct, integrated, Freq difference data of ANY noise type and does it without adding any dead time or aliasing all by using pretty much using ANY ADC capability of over sampling at the tau Zero rate. If one then uses the data log with something like the classic Stable 32 S/W or Ulrich's Plotter, it gives is the exact same results as other methods costing much much more, over the whole tau range. This is limited only be its reference oscillator (Same way that all others are limited of course, Doesn't get much better than that). If that is not good enough for you, them you need to discuss the results with Symmetricon and others that give the same answer as mine, not me. If for some reason you want to set one up wrong so that it matches the results of some other special instrument, I'd be glad to tell you how to have it add back in the dead time or aliasing artifact problems or whatever else you would like it to do wrong, that it presently does correctly. ws ** Bruce wrote As long as one is aware that your method (as implemented by you) doesn't actually measure Allan variance, it may be useful for comparing the relative stability some sources for small Tau (unfortunately the range of Tau for which the method may produce useful results depends on the phase noise characteristics of the sources being compared). To measure AVAR the technique has to have the same response to all phase noise spectral components as does AVAR. Since you do not integrate/average the frequency measures the phase noise response of the method is not identical to that used in calculating AVAR. This technique probably works best when white phase noise dominates the phase noise spectral region of interest (usually for small Tau). For
Re: [time-nuts] Dual Mixer
Spreadsheets can be a snare and a delusion if not carefully applied with full recognition of their limitations. Unfortunately the divergent nature of flicker phase noise etc doesn't become evident until one processes a very large number of samples. 1000 samples is never enough as many tests and simulations both published and unpublished have shown. Your assertion about scale factors is questionable as the equivalent noise bandwidth needs to be identical when comparing measures produced by different systems. If they differ the results will differ by a scale factor. Aliasing will increase the equivalent noise bandwidth somewhat so exact matching may be difficult. Also its not possible to reconstruct the necessary samples that would be produced by an ideal integrator using a spreadsheet from a finite sequence of samples that the spreadsheet can handle as the number of filter coefficients required is too large for the spreadsheet to cope. WarrenS wrote: Bruce So why are you saying I need millions of samples? Is it that this method of integration may give the wrong answer one out a million times? And you will not let up until you find that one in a million times that it may error? I don't think you're going to find it, but if you want we can go with that. BTW It does NOT need ANY scale factors, special or otherwise to give the right answers. It uses the same scale factor of ONE for ALL noise sources. If you can't give me an example of a data log that it may fail on, that I can run thru excel to prove otherwise, then We're done here until next time. ws *** * The results have so far only been shown to be useful when white phase noise dominates. When the phase noise is white almost anything can be made to produce a result that differs from ADEV by at scale factor. In practice its sometimes difficult to know over what range of Tau that the phase noise is in fact white. The various tests and comparisons that have been made or are underway are necessary but not sufficient proof of the usefulness of this technique. The phase noise frequency response of the technique is also required so that its limitations can be delineated. 1000 samples of a divergent noise process are insufficient, spreadsheet analysis of the millions of samples that are probably necessary is impossible/impractical. Using something like Matlab is probably necessary to achieve meaningful results. Bruce *** WarrenS wrote: OK, So, It is not perfect, but its simple and does give answers that are GOOD enough. At least you now understand if the integrator works then the tester works. So that we do not go down hill again after all this progress, If you would like to send me a data file of say 1000 + samples of any noise type of your choice I'll send you back an excel spread sheet to show the insignificant error that this integration method produces. ws You cannot approximate the sinc function frequency response of an ideal integrator with an arbitrary low pass filter. Your scheme will tend to misbehave (in that it will produce anomalous ADEV estimates) when flicker phase noise is significant. You actually need to use an analog low pass filter (or its equivalent) and an integrator to produce useful ADEV measures Bruce *** WarrenS wrote: (My apologies to all, this is a game Bruce and I play every time I bring up my simple tester.) Bruce wrote: So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Yes, I integrate/average just the same as I have always done it from day one. Did you finally understand how the integration works using most any ADC? Hint: it's done with oversampling the tau zero time. (and a LP filter set to a value above the tau zero but below the oversamping rate) The VERY SAME thing I have been trying to tell you from day one, something that you have chosen to ignore. The very original Block diagram that I posted shows it, if you need more information. ws *** Warren So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Bruce * WarrenS wrote: Bruce Before we go around again and discuses what my simple tester can and can not do and why, It would be helpful if you would take the time to better understand how it works and why it works the way I have done it. You really should try one yourself if you can't see why it works. You are going to be surprised and embarrassed at how good it works. Why you're at it, try the swing test with anything you have. Let me know how that goes. I'm not saying that may tester will match someone's Latest ever changing NEW idea of what the correct AVAR should be, After all it just Logs correct, integrated, Freq difference data of ANY noise type and does it without adding any dead time or aliasing all by
Re: [time-nuts] Dual Mixer
Warren Calculating an integral using a sampled data system when the Nyquist criterion is met is exactly equivalent to filtering albeit using just the right coefficients. Using rectangular approximation to the integral of the underlying continuous function is also equivalent to a filter albeit a very simple one. Unfortunately rectangular integration (which you use) isnt particularly accurate, using trapezoidal integration is far more accurate in most cases. Since this isnt a control system the instability associated with trapezoidal integration and higher order integration algorithms in feedback systems isnt an issue. Whittaker-Shannon-Kotelnikov interpolation allows an exact reconstruction (when the Nyquist sampling criterion is met) of the underlying continuous function from the samples. The result can then be integrated term by term to produce a set of weights/filter coefficients for the data samples. In other words in a sampled data system integration is equivalent to using a filter. Near enough is never good enough if you cant estimate the errors involved in the various approximations. This is particularly true when one is attempting to evaluate the deviation of an approximate method from that achieved using the correct method. Bruce WarrenS wrote: Bruce So why are you saying I need millions of samples? Is it that this method of integration may give the wrong answer one out a million times? And you will not let up until you find that one in a million times that it may error? I don't think you're going to find it, but if you want we can go with that. BTW It does NOT need ANY scale factors, special or otherwise to give the right answers. It uses the same scale factor of ONE for ALL noise sources. If you can't give me an example of a data log that it may fail on, that I can run thru excel to prove otherwise, then We're done here until next time. ws *** * The results have so far only been shown to be useful when white phase noise dominates. When the phase noise is white almost anything can be made to produce a result that differs from ADEV by at scale factor. In practice its sometimes difficult to know over what range of Tau that the phase noise is in fact white. The various tests and comparisons that have been made or are underway are necessary but not sufficient proof of the usefulness of this technique. The phase noise frequency response of the technique is also required so that its limitations can be delineated. 1000 samples of a divergent noise process are insufficient, spreadsheet analysis of the millions of samples that are probably necessary is impossible/impractical. Using something like Matlab is probably necessary to achieve meaningful results. Bruce *** WarrenS wrote: OK, So, It is not perfect, but its simple and does give answers that are GOOD enough. At least you now understand if the integrator works then the tester works. So that we do not go down hill again after all this progress, If you would like to send me a data file of say 1000 + samples of any noise type of your choice I'll send you back an excel spread sheet to show the insignificant error that this integration method produces. ws You cannot approximate the sinc function frequency response of an ideal integrator with an arbitrary low pass filter. Your scheme will tend to misbehave (in that it will produce anomalous ADEV estimates) when flicker phase noise is significant. You actually need to use an analog low pass filter (or its equivalent) and an integrator to produce useful ADEV measures Bruce *** WarrenS wrote: (My apologies to all, this is a game Bruce and I play every time I bring up my simple tester.) Bruce wrote: So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Yes, I integrate/average just the same as I have always done it from day one. Did you finally understand how the integration works using most any ADC? Hint: it's done with oversampling the tau zero time. (and a LP filter set to a value above the tau zero but below the oversamping rate) The VERY SAME thing I have been trying to tell you from day one, something that you have chosen to ignore. The very original Block diagram that I posted shows it, if you need more information. ws *** Warren So you now actually integrate/average the frequency over the sampling interval (Tau) after rejecting the need to do this for months? Bruce * WarrenS wrote: Bruce Before we go around again and discuses what my simple tester can and can not do and why, It would be helpful if you would take the time to better understand how it works and why it works the way I have done it. You really should try one yourself if you can't see why it works. You are going to be surprised and embarrassed at how good it works. Why you're at it, try the swing test
Re: [time-nuts] oscillator choice question
Hal Murray wrote: bruce.griffi...@xtra.co.nz said: If there is no electronic tuning available one can use a DDS based synthesiser to produce a corrected output frequency. However close in spurs will be problematic unless one use a couple of simple mix and divide stages or resorts to a Diophantine synthesiser using phase noise truncation spur free output frequencies from the DDS chip(s). I think I understand the classic spurs from a DDS. I wasn't familiar with Diophantine techniques. Google found this http://www.hindawi.com/journals/ijno/2008/416958.html which is readable at my level. But I don't think I understand the big picture. The example numbers they give involve mixing 500 Hz with 10 MHz. Assuming I want the sum, how do I get rid of the difference? It's going to be a good strong signal, as strong as the one I want. I think anything that leaks through the filter into the next mixer is likely to make mirror sidebands that are right where we don't want them. Why is that going to be easier to get rid of than traditional spurs? A DDS can generate some close in spurs that are very close to the desired frequency and thus are difficult to filter even with a narrow band PLL as the spur offset and amplitude varies with the DDS output frequency in a very complex way. A very narrow PLL requires a VCO with good short (for averaging times up to the inverse PLL bandwidth) term stability. However if the offset is 500Hz its relatively easy to filter out the unwanted sum (or difference) frequency with a PLL using a VCO with good short term stability for averaging times of a few tens of millisec. The rejection can be improved by using an SSB mixer. N.B. the author of the paper (and his web page) that you found has vanished without trace. It turns out that the Diophantine frequency synthesis technique was patented (US Patent 5267182) some 17 years ago. His literature search for previous papers/patents cant have been very effective/extensive. Fortunately I managed to download all of his papers before they vanished along with the web page. Alternatively if one implements the DDS in an FPGA its possible to virtually eliminate such spurs using a modified algorithm. However this requires an external DAC to produce the required output. Got a URL? What's magic about a FPGA? Why don't traditional DDS chips use that modified algorithm? http://www.sdrforum.org/pages/sdr06/sdr06_papers/1.3/1.3-01.pdf (thanks to Bob Camp for finding this gem) There's nothing magic about an FPGA, its merely a convenient way of implementing the improved algorithm. There's no way to implement it with a traditional DDS chip as the digital section needs to be extensively modified. DDS chips do not use it because it has only recently been devised. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] oscillator choice question
Bob Camp wrote: Hi A coupe of issues with mechanical servo tuning: 1) It wears out the tuning capacitor pretty fast. They are designed for a limited number of adjustments. They loosen up with a lot of tuning and this degrades their stability. 2) It would be much easier to tear apart the mechanical tune OCXO and put in a tuning diode than to rig a thermally isolated high resolution servo stepper 3) Mechanical tune arrangements normally have backlash. That's not an issue as long as the servo only goes one way. It becomes a real pain to correct for each time you reverse direction. One solution to which is to add (in addition to the servo motor) a torque motor to preload the gear train so that the same flank of each gear tooth is in contact for both directions of rotation. Zero backlash drive reduction systems are also available at some considerable cost. 4) Making a mechanical setup with a minimum step below 1 ppt is going to be more than just a simple stepper. A gear chain based system will be pretty exciting to work up. Backlash in the gears will add to what ever you have in the tune it's self. 5) The tuning on the OCXO may not be monotonic. That's especially true if you do indeed run the trimmer at a higher resolution than a normal human could adjust it. Tuning reversals tend to drive servo loops a bit crazy. None of that says that it can't be done. All it says is that it will be hard to do well. What kind of accuracy are you trying to obtain? Bob Bruce On May 2, 2010, at 3:18 PM, ch...@yipyap.com wrote: That is a really cool picture. Can I be like you when I grow up? I've figured out which of these silvered modules in this Schomandl sig gen is the oscillator (the one that got warm). I have to figure out if it is voltage adjustable in some way. Does anyone use mechanical adjustment with a servo, gear train and microcontroller? I did hear all of those good advisers telling me to buy the Thunderbolt. But I already have these pieces so... -- Chris w0ep Niels Lueddecke wrote: You see? Don't do it, don't even think about starting. Go buy a trimble thunderbolt, it will save you LOTS of time! http://www.dulli.org/pics/20100502%20-%20Clock.jpg ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] oscillator choice question
If there is no electronic tuning available one can use a DDS based synthesiser to produce a corrected output frequency. However close in spurs will be problematic unless one use a couple of simple mix and divide stages or resorts to a Diophantine synthesiser using phase noise truncation spur free output frequencies from the DDS chip(s). Alternatively if one implements the DDS in an FPGA its possible to virtually eliminate such spurs using a modified algorithm. However this requires an external DAC to produce the required output. Bruce Robert Atkinson wrote: Hi Chris,The Racal high stability units usually use the 9420 series OCXO's. These are good oscillators but do not have electronic tuning as standard. 'they are also normally 5MHz. What is the best oscillaor depends on your requirements. The two main parameters are phase noise and hold-over performance. Hold over is how much the oscillator will drift if the GPS loses signal. Robert G8RPI. --- On Sat, 1/5/10, ch...@yipyap.comch...@yipyap.com wrote: From: ch...@yipyap.comch...@yipyap.com Subject: [time-nuts] oscillator choice question To: time-nuts@febo.com Date: Saturday, 1 May, 2010, 20:28 I'd like to build a GPS disciplined frequency standard. I am slowly gathering up pieces. I have a Trimble Resolution T GPS card that appears to work, and an antenna for it. I'm thinking now of the oscillator part. I have two Racal 1992 counters with the stable oscillator option (probably 04E since these are former military units). I also have an old Schomandl ND-100M Frequenzdecade signal source with an (I assume) ovenized oscillator w/unknown properties. I'm wondering if I could use an oscillator from one of these gizmos instead of shelling out real money on ebay? Speaking of which, it seems like half the people in China are selling oscillators. I assume some of them are good for this application and some not so good? The usual suspects from HP and Agilent are there, and they seem to command a pretty good (high) price. Which is why I'm eyeing the Schomandl. Chris w0ep ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Frequency Stability of An Individual Oscillator:Negative Values?
There has been some work done on the effect of finite correlation between oscillator outputs. In some cases allowing a finite correlation coefficient improves the N cornered hat ADEV estimates. In most cases the oscillators being compared share the same ambient environment and thus may exhibit correlated fluctaution due to ambient paramaeter (temperature, pressure humidity etc) variations. Bruce Bob Camp wrote: Hi How to treat a negative is up to you, it's obviously indicating a not real outcome. Zero is also a not real for realizable oscillators. Most simply note the result as below floor, drop it, and proceed. Since the variability of the data is driving the negative results, it's unlikely that another approach will massively improve things (with the same data set). The practical answer is to use oscillators with closer noise performance to reduce the scatter or to improve the data collection method if it's the limiting factor. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Kyle Wesson Sent: Thursday, April 22, 2010 5:07 PM To: time-nuts@febo.com Subject: [time-nuts] Frequency Stability of An Individual Oscillator:Negative Values? Hello, I am working to determine the Allan variance of an individual oscillator from a series of three paired measurements as described in the paper by Gray and Allan A Method for Estimating the Frequency Stability of An Individual Oscillator (NIST, 1974, tf.nist.gov/general/pdf/57.pdf). In this report they make reference to the statistical uncertainty of the measurement due to ensemble noise and potential clock phase correlation which can potentially make the Allan variance for an individual oscillator have a negative value. They write: If the noise level of the oscillator being measured is low enough, and the scatter high enough, equation (4) may occasionally give a negative value for the variance. My question is: how should I treat negative variance values in this case? For example, if my data set were to produce an individual oscillator Allan variance with a value of -5e-12, should I convert this value to 0 (ie. the closest valid sigma value to the number since 0= sigma inf ), take the absolute value of the result (ie. turn -5e-12 to +5e-12), or drop the result from my estimate of individual oscillator frequency stability altogether? Is there another method that will produce estimates of individual oscillators from an ensemble approach but assures non-negative output variances? Thank you in advance, Kyle ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] FEI FE-5680A
Leigh L. Klotz, Jr WA5ZNU wrote: On 03/24/2010 03:18 PM, Magnus Danielson wrote: Fellow time-nuts, ... Second, anything I should keep in mind as I power one up? ... Cheers, Magnus There's a bit of discussion in the archives about the need for a heat sink, and also about the whether it's necessary to anneal the case in a 400C Hydrogen reducing oven after the SMA modification. (Although I do have access to one, it is over the threshold for things I'm willing to do.) Leigh. If one were to use a trepaning tool with loose abrasive and plenty of water coolant and slowly grinds the required hole through the mu metal cover, the thermal stress and mechanical disturbance of the mu-metal shield should be minimised precluding the need for hydrogen annealing. This is easily done using a drill press and a suitable tool. The tool can be assembled from brass tubing and rod. Since the forces involved are low even soft soldering should suffice. Water jet cutting should also work well. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS Display clock
b...@lysator.liu.se wrote: Zealand is an island between Denmark and Sweden. Was not aware of that. Lat/Lon? Zeeland is a province of the Netherlands Been there... Bruce -- Björn The name New Zealand originates from the latter via Dutch Cartographers. Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] HP10514B Mixers
Upload it to the manuals section of Didier's site: http://www.ko4bb.com/cgi-bin/manuals.pl Bruce Brian Kirby wrote: The manual is about a megabyte - the list will not accept it. ewkeh...@aol.com wrote: Is there a way you can post it on time-nuts? Thank you Bert Kehren In a message dated 3/30/2010 9:10:19 P.M. Eastern Daylight Time, kilodelta4foxm...@gmail.com writes: We'll the As are the package with the BNC connector. The Bs are for OEM equipment, they had 6 pins and a smaller plastic black case Turns out they have some of the lowest phase noise specs on record I can send the manual, if you like... Brian Mike Feher wrote: I do not believe that I am familiar with the 10514B. I probably have about 6 or so 10514As with BNC connectors (most in a metal case and a few in epoxy), and, I recall from about 40 years ago a 10514A that was a small little black cube with leads for mounting on a PC board. Naturally, I had to take one apart back then as that was the fun part. In the late 60's I designed a thick-film phase detector that was used in an ASW system, where I simply matched the hot carrier HP diodes for a fixed voltage drop at a given current. We also wound the toroids, and had an op-amp on the output. Wow, that was ages ago. Regards - Mike Mike B. Feher, N4FS 89 Arnold Blvd. Howell, NJ, 07731 732-886-5960 ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.