Re: [time-nuts] 5071A Cs Oven
Paul- Same here! :- ) The best way is to find a REALLY dead tube and cut it open and see if the EM first and second dynodes are clean, or covered in a film of Cs. Short of heating the dynode to vaporize off the Cs and let the ion pump collect it, I think it becomes a giant physics project. One really crazy idea is to cut the steel casing open, but leave the glass seal intact. If that's even possible. Then. you could use RF inductive heating like what is used when you make a classic vaccum tube, to heat the elements inside and vaporize the occluded (sp?) gases from the elements and let the getter material or the ion pump gather the crud up. My only guess is if that was a real practical way of adding life back to a Cs tube, that HP or somebody would have been doing it in the 1970's when vaccum tubes were still in mass production and the effort would have paid off. In light of the fact that Symetricom only wants a paultry $18k for a new CBT... and now with GPS birds running Rb's rather than CsAND you can get a CSAC for $1500 new, the demand for Cs tubes has but a limited market. Now if I could convince my XYL that I must spend $250k on an H-MASER then I'd be in Time-Nut's Nerdvana! -Brian, WA1ZMS On Oct 4, 2012, at 4:59 PM, paul swed paulsw...@gmail.com wrote: Brian Funny you mention the dirty emult. At the time I was wondering and if there was any insane way to get the Cs off the e multiplier. HV to ground or something to migrate them. Hold upside down and shake hard. ;-) Regards Paul On Thu, Oct 4, 2012 at 4:22 PM, Brian, WA1ZMS wa1...@att.net wrote: Rick can correct me on this... but my understanding is that it is almost never a case of running out of Cs from the oven, but that the dynodes of the electron multiplier become covered in used Cs and so the ultimate SNR and effective beam current falls to a point where there is no dependable beam current left to do much with. The tube that I sent to Paul, he was able to get some more current by running the Cs oven at a higher temp and as such is vaporizing off LOTS of Cs and can get a tiny bit of beam current to work with. But it is a down hill slide, since to keep the beam current up with a dirty E-mult, over time you need to keep running up the Cs oven temp to the point where you may just run out of Csor get an E-mult that is so covered with used Cs that there is no beam current to work with. You can always try and run up the E-mult voltage as well, but you then run the risk of arcing in the E-mult and an issue of burning up the resistor divider stack that I believe is part of the E-mult. I can see it now.. an end-of-life Cs tube that needs 35kV on the E-mult to get any beam current to work with! : -) -Brian, WA1ZMS On Oct 4, 2012, at 3:41 PM, paul swed paulsw...@gmail.com wrote: Rick Well I have the old 5060 and 61 tubes and indeed they run out O gas. I am using a 5060 tube in a 5061 and the Frankenstein works though it has only a few Cs. But heck it all locks after 48 hours. So someone said with humor lets buy some 5071 tubes for a couple hundred each. Guess I could have some fun getting that into the 5061. Suspect that would be the bride of Frankenstein then. Thanks for sharing the insights and experience. Paul WB8TSL On Thu, Oct 4, 2012 at 3:21 PM, Richard (Rick) Karlquist rich...@karlquist.com wrote: These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On
Re: [time-nuts] Best counter setting for ADEV?
Hi Doing the repetition may give you a more fine grained idea of what the clocks are doing. It does not impact the resolution of the single measurement. A single ns in 100,000 seconds is indeed 1.0x10^-14… Bob On Oct 4, 2012, at 6:11 PM, Azelio Boriani azelio.bori...@screen.it wrote: OK, because you think: (Y-X)/100K is the drift of the two clocks for each second over 100K seconds and for each second I have a 1nS/100K resolution. Now think about this: repeat the measure over and over and after each 100K seconds you have your table. This table (your counter is 1nS) can have figures like, say, 52nS, 53nS, 58nS, that is with a 1nS step. Your drift maybe not exactly 52: it maybe 52.1 52.3 52.4 but you only see 52. So you are loosing an entire nS between 52nS and 53nS. If you have a 100pS counter you can see 52.0, 52.1 and so on. On Thu, Oct 4, 2012 at 10:13 PM, Bob Camp li...@rtty.us wrote: Hi If I have two clocks: 1) I know they are X ns apart at time = 0 2) I know they are Y ns apart at time = 100,000 seconds 3) The resolution of the X and Y measurements is 1 ns I know the relative time between the two clocks to a lot better than 5.0x10^10... Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 3:43 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? If you take 2 samples out of a 1nS counter than you can estimate your measure with a 500pS resolution (5E-10) with an uncertainty (noise) of 7E-10. On Thu, Oct 4, 2012 at 8:58 PM, Bob Camp li...@rtty.us wrote: Hi What if I only take two single measurements: One at time = 0, the other at time = 100,000 seconds. No averaging, no signal processing just two measurements. I look at the time difference between the two signals at time = 0, and then again a bit more than a day later. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 2:48 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? You're right: it is better to put it down correctly: for a theoretical resolution of 10E-14@100K seconds, start with a 1nS counter that takes 1 second samples for 100K seconds and average those 100K samples. You have your resolution and a noise (an error bar) of 3E-12. If you use a 100pS counter and do the same run, you will end up with a 10E-15 reolution and 3E-13 noise, and so on. It turns out that to have a real 10E-14 measure (@100K seconds), not blurred with noise, you must start with a 10pS counter. On Thu, Oct 4, 2012 at 6:41 PM, Bob Camp li...@rtty.us wrote: Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the
Re: [time-nuts] 5071A Cs oven
I bought a couple of 5071As on eBay recently, on the basis that the various status messages shown in the auctions didn't look like tube failures. They both reported Cs oven voltages at 0.0. The first unit turned out to have a severe intermittent noise problem, originating somewhere other than the tube or OCXO, that kept the unit from being able to pass its own power-up self test most of the time. This one had a 10890A standard-performance tube. The other 5071A was in much better physical shape. It had a 10891A high-perf tube that looked practically brand new, with almost no dust on the HV leads. The power-up self test also failed on this one, but for a different reason: the tube had no beam current at all. I was able to get the second 5071A working perfectly with the standard 10890A tube from the other one. At that point, I compared the resistance and capacitance-to-ground readings between the two tubes' connectors to try to figure out why the 10891A wasn't working. The cause of the failure wasn't immediately clear. The resistance readings were comparable on the oven heater, C field coil, and thermistor, nothing was shorted, and the vacuum was good. It passed all of the beam-tube tests in the 5071A firmware. The only odd thing I noticed about the bad tube, besides the lack of beam current from its output dynode, was the fact that its output was rather microphonic, in a triboelectric sense. Tapping on the bad tube's housing with a screwdriver yields a nice ringdown voltage even with no power applied to the tube (see http://www.ke5fx.com/cs_ring.png ). I could even see something resembling a speech waveform if I yelled at the tube (and believe me, I tried that.) With the good tube, there was no trace whatsoever of this microphonic behavior. So, this one has me stumped. It's really frustrating because I've got to assume that I've got a high-perf 5071A tube in near-mint condition that would be a superb performer if it weren't for an apparently-simple internal defect. Perhaps it was dropped, breaking loose something in the EM assembly that is now rubbing against a dissimilar material? I'm going to save this tube in case I ever have a vacuum rig of some kind that would allow laparoscopic surgery to be performed on it. Or perhaps I'll have a chance to X-ray it at some point. -- john, KE5FX Miles Design LLC Cs depletion in modern tubes is a real failure mode! However some competing modes are: Ion pump failure (whiskers or high impedance shorts) High Cs backround level (Gettering saturation) Oven or ionizer filament failure (More common on older HP units with the AC excitation, (think of those neat light bulbs with the magnet inside where the filament jumps all over the place, less drastic mechanical flexing caused the filaments to eventually fail due to fatigue.) Slow leaks ruining the vacuum Running the Cs with the ovens off as described WILL delay some of these failures, (it's recommended that the ion pump remain energized however) A 5071A tube could be installed into a 5061A (with proper adaptors) but why? Later HP 5061A and 5061B mini tubes with the same diameter as the 5071A tubes can be installed into the 5071A with some caveats and proper adaptors. Corby ___ 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] Tracking NTP displacement and correlation between two clients.
David Taylor has all sorts of NTP monitoring scripts, software, and tips at his web site. Start at http://www.satsignal.eu/software/net.htm#NTPmonitor and look around. Brent Thanks for the mention, Brent. Two Windows-based programs: Remote comparison of multiple NTP servers: http://www.satsignal.eu/software/net.htm#NTPmonitor Plotting and analysis of NTP statistics files: http://www.satsignal.eu/software/net.htm#NTPplotter Sample plots using MRTG: http://www.satsignal.eu/mrtg/performance_ntp.php Perhaps some of those tools may help Bob or others on the list. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-tay...@blueyonder.co.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.
Re: [time-nuts] Tracking NTP displacement and correlation betweentwo clients.
The problem stems from one of the two (identical) machines drifting off by 60-70 seconds per day. So a few ms here and there are ok. [] Bob == Bob, NTP is normally limited to a +/- 500 parts per million correction - 43 seconds per day. You may be operating outside the range NTP is expecting to handle. If the clock offset is a stable value of 60-70 seconds per day you can bias NTP to correct within +/- 500 ppm of that drift. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-tay...@blueyonder.co.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.
Re: [time-nuts] Tracking NTP displacement and correlation betweentwo clients.
David, The problem is that they start in sync and over the course of a day drift that far apart despite having NTP running. We're not sure why NTP isn't correcting it along the way. Though at this point, we are looking at a firmware bug. Thanks! Bob On Oct 5, 2012, at 12:30 AM, David J Taylor david-tay...@blueyonder.co.uk wrote: The problem stems from one of the two (identical) machines drifting off by 60-70 seconds per day. So a few ms here and there are ok. [] Bob == Bob, NTP is normally limited to a +/- 500 parts per million correction - 43 seconds per day. You may be operating outside the range NTP is expecting to handle. If the clock offset is a stable value of 60-70 seconds per day you can bias NTP to correct within +/- 500 ppm of that drift. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-tay...@blueyonder.co.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] Tracking NTP displacement and correlation betweentwo clients.
I had an early Phenom II that lost time while turned on but the internal CMOS clock did not so rebooting or reading the CMOS clock restored the correct time. There was a problem with the System Management Mode code and The C1E CPU state which was new at that time where an interrupt was being lost. Disabling the low power CPU state fixed it until a BIOS update was released. It has been a while but as I recall, the NTP client kept the OS from drifting further behind but the time was still noticeably off. On Fri, 5 Oct 2012 00:42:42 -0400, Bob Bownes bow...@gmail.com wrote: David, The problem is that they start in sync and over the course of a day drift that far apart despite having NTP running. We're not sure why NTP isn't correcting it along the way. Though at this point, we are looking at a firmware bug. Thanks! Bob On Oct 5, 2012, at 12:30 AM, David J Taylor david-tay...@blueyonder.co.uk wrote: The problem stems from one of the two (identical) machines drifting off by 60-70 seconds per day. So a few ms here and there are ok. [] Bob == Bob, NTP is normally limited to a +/- 500 parts per million correction - 43 seconds per day. You may be operating outside the range NTP is expecting to handle. If the clock offset is a stable value of 60-70 seconds per day you can bias NTP to correct within +/- 500 ppm of that drift. Cheers, David -- SatSignal Software - Quality software written to your requirements Web: http://www.satsignal.eu Email: david-tay...@blueyonder.co.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. ___ 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] Best counter setting for ADEV?
Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] Best counter setting for ADEV?
Azelio, as an example, with a 53131A you have an output 'granularity' of +/- 500 ps. No matter how much averaging happens inside the counter, if the least significant digit is 500 ps, then the (ADEV) measurement limit is always 5E-10 at 1 sec as I posted in the initial post to this thread. This is not the accuracy of the counter, it's simply the available display and output string resolution. At 10 sec., the ADEV limit of the same counter is 5E-11 etc. etc. finally getting limited by the reference source and other effects. If the DUT's ADEV is in the 1E-13's, any measurement values below 1000 sec. are void and represent just the measurement limit of the +/- 500ps counter. Now I'm waiting to receive 'glasses' for my 53131A that are in transit :) Adrian Azelio Boriani schrieb: Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] Best counter setting for ADEV?
OK, and then what about measurements beyond 1000 sec (for the same DUT's ADEV in the 1E-13's) taken with a 500pS counter and a 50pS counter? Nice to hear that there is exactly no difference, just use average... On Thu, Oct 4, 2012 at 2:34 PM, Adrian rfn...@arcor.de wrote: Azelio, as an example, with a 53131A you have an output 'granularity' of +/- 500 ps. No matter how much averaging happens inside the counter, if the least significant digit is 500 ps, then the (ADEV) measurement limit is always 5E-10 at 1 sec as I posted in the initial post to this thread. This is not the accuracy of the counter, it's simply the available display and output string resolution. At 10 sec., the ADEV limit of the same counter is 5E-11 etc. etc. finally getting limited by the reference source and other effects. If the DUT's ADEV is in the 1E-13's, any measurement values below 1000 sec. are void and represent just the measurement limit of the +/- 500ps counter. Now I'm waiting to receive 'glasses' for my 53131A that are in transit :) Adrian Azelio Boriani schrieb: Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] There are a couple CSACs for sale
on that big internet auction site i don't know by who etc, etc. -pete ___ 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] Best counter setting for ADEV?
Hi But I don't *have* to take data only at one second points for a one second ADEV. I can take data faster and then process it. That's what the counter does in frequency mode. I can do the same thing with time readings. The big advantage there is that with higher speed time readings, I can control how the math works and *hopefully* get a valid ADEV out the other side. With the counter doing it's magic math in the frequency mode, you are entirely at the mercy of an unknown HP coder who might or might have been thinking ADEV when he wrote the code. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Adrian Sent: Thursday, October 04, 2012 8:35 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Azelio, as an example, with a 53131A you have an output 'granularity' of +/- 500 ps. No matter how much averaging happens inside the counter, if the least significant digit is 500 ps, then the (ADEV) measurement limit is always 5E-10 at 1 sec as I posted in the initial post to this thread. This is not the accuracy of the counter, it's simply the available display and output string resolution. At 10 sec., the ADEV limit of the same counter is 5E-11 etc. etc. finally getting limited by the reference source and other effects. If the DUT's ADEV is in the 1E-13's, any measurement values below 1000 sec. are void and represent just the measurement limit of the +/- 500ps counter. Now I'm waiting to receive 'glasses' for my 53131A that are in transit :) Adrian Azelio Boriani schrieb: Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] Best counter setting for ADEV?
Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] Tracking NTP displacement and correlation between two clients.
It had to happen eventually. Time Nut interest overlapped with $DAY_JOB. Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. Thanks! 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.
Re: [time-nuts] Tracking NTP displacement and correlation between two clients.
Bob, check http://www.febo.com/pages/plots/ntp and see if the graphs there would provide the info you need. They're based on using one NTP server to monitor the offset of a number of other services and plotting the results over time. If so, I'm happy to make the (*nix-based) scripts available. John On 10/4/2012 10:44 AM, bownes wrote: It had to happen eventually. Time Nut interest overlapped with $DAY_JOB. Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. Thanks! 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] 5071A Cs Oven
Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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.
Re: [time-nuts] There are a couple CSACs for sale
Just sold. It was not me: they was shipping USA-only. Very interesting price, anyway they are just gone. On Thu, Oct 4, 2012 at 4:05 PM, Pete Lancashire p...@petelancashire.comwrote: on that big internet auction site i don't know by who etc, etc. -pete ___ 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] Tracking NTP displacement and correlation between two clients.
On Thu, 4 Oct 2012 10:44:41 -0400 bownes bow...@gmail.com wrote: I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. I guess, the simplest method would be to add system A to the ntp server list of system B, but as a non-refernece station (don't remember the flag needed for that, but it's somewhere in the documentation), then you can track the difference between the two systems by using the peers log on system B. Of course, it will differ from what the value actually is, but it should be accurate within one RTT between system A and B. If you cannot track one system from the other, using a different NTP server that can be reached from both, could do the job. Again, let the system A and B track it as non-reference station, but make sure to keep a high poll rate (the standard 1024s is too low and will not get you enough data to do usefull statistics with short measurement times). Of course you can also use a GPS as reference ntp server on system A and system B and measure the offset against that. Doing a plot from the peers log is then a simple matter of handling gnuplot and maybe a line of perl for statistical analysis (or excel if you prefer that other OS ;-) That's the proof of the measurement style. The other possibility would be to have both peer logs of system A and B analysed for the RTT and jitter values. From that you can argue that the time difference between the two systems is at most (RTT_A_max+RTT_B_max)/2 (theoretically one leg of the path to the ntp server could have zero delay. thus the calculated delay would be off by half the RTT). This will clearly overestimate the time difference between the two systems. A more accurate time difference value would be jitter_A+jitter_B under the assumption that the packet delays on both legs of the path are nearly identical[1]. There is no division by 2 using the jitter, because jitter does not need to be symetrical around an average, but can have any probability distribution. Thus (jitter_A+jitter_B)/2 could underestimate the actuall time difference. [1] this assumption should hold true for most networks these days given that: 1) the jitter is much smaller than RTT 2) none of the network connections is conguested You can test it if you do a traceroute from both ends of the path. If both tracerouts show the same routers, then the path that the packet travels on the its way back and forth is the same. HTH Attila Kinali -- There is no secret ingredient -- Po, Kung Fu Panda ___ 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] Best counter setting for ADEV?
Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] Best counter setting for ADEV?
Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] 5071A Cs Oven
Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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.
Re: [time-nuts] Tracking NTP displacement and correlation between two clients.
bow...@gmail.com said: Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. Assuming you are running the standard ntpd... It includes all sorts of logging. Set up the two systems so they use each other as servers. Turn on rawstats. ntpd will add a line each time it exchanges a pair of packets with a server. That line will have the IP Address and 4 time stamps. See the documentation. Details are in monopt.html The 4 time stamps are: time the request left the local system time the request arrived at the remote system time the response left the remote system time the response arrived at the local system If you subtract the first two, you get the network transit time for the request packet as skewed by the clock offset. Subtracting the last two gives you the transit time for the response packet. If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. (If you assume the clocks are both accurate, you can compute the network transit time in each direction.) If you want to graph the results, you have to split out the lines for the server you are interested in. Then you can feed it to gnuplot/whatever. You can also do the monitoring from another system, but then you have to sort out what happens when the clock on that system is off. How good is your connection to the big bad internet? If you run a big download over a slow link, the queuing delays can confuse ntp. You might want to look at the timings from your systems to the stratum-2 servers and/or from the stratum-2 servers to the outside world. -- These are my opinions. I hate spam. ___ 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] Tracking NTP displacement and correlation between two clients.
On Thu, 04 Oct 2012 09:46:38 -0700 Hal Murray hmur...@megapathdsl.net wrote: If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. Within a LAN, RTT is usually in the range of 200us with a jitter in the same range (it can happen that jitter is significanlty higher than RTT, if you have bursty traffic in your LAN) Attila Kinali -- There is no secret ingredient -- Po, Kung Fu Panda ___ 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] 5071A Cs Oven
Hi The associated log entries are: Ion pump over current Cs oven failed Not encouraging... Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Tom Van Baak Sent: Thursday, October 04, 2012 12:45 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] 5071A Cs Oven Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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.
Re: [time-nuts] Tracking NTP displacement and correlation between two clients.
On Thu, Oct 4, 2012 at 12:46 PM, Hal Murray hmur...@megapathdsl.net wrote: bow...@gmail.com said: Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. redacted Assuming you are running the standard ntpd... It includes all sorts of logging. Set up the two systems so they use each other as servers. Turn on rawstats. ntpd will add a line each time it exchanges a pair of packets with a server. That line will have the IP Address and 4 time stamps. See the documentation. Details are in monopt.html The 4 time stamps are: time the request left the local system time the request arrived at the remote system time the response left the remote system time the response arrived at the local system That looks like a bit of overkill. :) If you subtract the first two, you get the network transit time for the request packet as skewed by the clock offset. Subtracting the last two gives you the transit time for the response packet. If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. In this case, the transit times should average out be very very close. The two machines in question are plugged into adjacent network ports with the same length of cable and the NTP server is on the same (lightly loaded) sub net. The problem stems from one of the two (identical) machines drifting off by 60-70 seconds per day. So a few ms here and there are ok. How good is your connection to the big bad internet? If you run a big download over a slow link, the queuing delays can confuse ntp. You might want to look at the timings from your systems to the stratum-2 servers and/or from the stratum-2 servers to the outside world. This is all relative to two internal stratum 2 servers on the same command and control network. No large xfers allowed over it. There are physically separate data, backup, and application networks for that. Thanks for the suggestions folks. I'm going to look into some of the standard ntp logging stuff as well as the scripts that John offered up. And now, I'm probably going to have set something up to start comparing my 3 GPSDO's and their associated machines! :) As an aside, I have seen that someone on the Raspberry Pi list has NTP running with 1pps into the GPIO ports. I've got a stack of GPS modules in stock with 1pps that are just itching to be tied to one of those, also in stock. Maybe this weekend. 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.
Re: [time-nuts] Tracking NTP displacement and correlation between two clients.
On Thu, 4 Oct 2012 13:04:27 -0400 Bob Bownes bow...@gmail.com wrote: The problem stems from one of the two (identical) machines drifting off by 60-70 seconds per day. So a few ms here and there are ok. Is it drifting without ntp or with ntp? 1minute drift per day is not unheard of for standard PC RTCs.. i've seen even worse.. This is all relative to two internal stratum 2 servers on the same command and control network. No large xfers allowed over it. There are physically separate data, backup, and application networks for that. On the average LAN (ie one that is connected by a couple of switches. No routers inbetween), the RTT will be dominated by the minimum packet times of ethernet. That's where the 200us delay comes from (actually it's below 1us for Gbit, but for some reason i've never seen it go under 100us). With modern switches that do worm-hole routing (ie just use the destination address in the first few bytes and then pass the packet on without further delay, while the end of the packet is still on the wire and not yet received by the switch) you dont even get much added delay from using multiple switches in line. With such large delays you can ignore any wire delay completely (which is in the range of 1us for maximum length cable). With routers it looks a bit different, there the whole packet is first stored in memory before being processed (due to more complicated routing decision). Due to this, RTT is still dominated by delay of the network hardware and not speed of light. Actually speed of light related delays will be buried deep in the noise unless you go trans-continental. And even then, router delay will still dominate. (eg RTT Europe-Japan is around 500ms, by which time a packet would be half way to the moon) So, if you have any significant time difference (1ms) between two systems that synchronize to the same NTP server in the same LAN, then the problem lies somewhere else than the network. Attila Kinali -- There is no secret ingredient -- Po, Kung Fu Panda ___ 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] Tracking NTP displacement and correlation between two clients.
David Taylor has all sorts of NTP monitoring scripts, software, and tips at his web site. Start at http://www.satsignal.eu/software/net.htm#NTPmonitor and look around. Brent On 10/4/2012 8:44 AM, bownes wrote: It had to happen eventually. Time Nut interest overlapped with $DAY_JOB. Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. Thanks! 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.
Re: [time-nuts] Tracking NTP displacement and correlation between two clients.
On Thu, Oct 4, 2012 at 1:31 PM, Attila Kinali att...@kinali.ch wrote: On Thu, 4 Oct 2012 13:04:27 -0400 Bob Bownes bow...@gmail.com wrote: The problem stems from one of the two (identical) machines drifting off by 60-70 seconds per day. So a few ms here and there are ok. Is it drifting without ntp or with ntp? 1minute drift per day is not unheard of for standard PC RTCs.. i've seen even worse.. With. Hence the issue. redacted So, if you have any significant time difference (1ms) between two systems that synchronize to the same NTP server in the same LAN, then the problem lies somewhere else than the network. Exactly where I am at this point. Dozen of other systems in the same facility don't seem to have this issue. ___ 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] 5071A Cs Oven
If this is the 5071A on that online auction site, I found out from the seller that the error log shows Cs oven timeout. This might be because of a bad oven supply (expensive) or a bad tube (really expensive)! Stan -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of time-nuts-requ...@febo.com Sent: Thursday, October 04, 2012 9:51 AM To: time-nuts@febo.com Subject: time-nuts Digest, Vol 99, Issue 30 Message: 6 Date: Thu, 4 Oct 2012 09:45:22 -0700 From: Tom Van Baak t...@leapsecond.com To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] 5071A Cs Oven Message-ID: CCE1197699D24CBDB6AC7D386B504875@pc52 Content-Type: text/plain; charset=iso-8859-1 Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob -- Message: 7 Date: Thu, 04 Oct 2012 09:46:38 -0700 From: Hal Murray hmur...@megapathdsl.net To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] Tracking NTP displacement and correlation between two clients. Message-ID: 20121004164638.c0fee800...@ip-64-139-1-69.sjc.megapath.net Content-Type: text/plain; charset=us-ascii bow...@gmail.com said: Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. Assuming you are running the standard ntpd... It includes all sorts of logging. Set up the two systems so they use each other as servers. Turn on rawstats. ntpd will add a line each time it exchanges a pair of packets with a server. That line will have the IP Address and 4 time stamps. See the documentation. Details are in monopt.html The 4 time stamps are: time the request left the local system time the request arrived at the remote system time the response left the remote system time the response arrived at the local system If you subtract the first two, you get the network transit time for the request packet as skewed by the clock offset. Subtracting the last two gives you the transit time for the response packet. If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. (If you assume the clocks are both accurate, you can compute the network transit time in each direction.) If you want to graph the results, you have to split out the lines for the server you are interested in. Then you can feed it to gnuplot/whatever. You can also do the monitoring from another system, but then you have to sort out what happens when the clock on that system is off. How good is your connection to the big bad internet? If you run a big download over a slow link, the queuing delays can confuse ntp. You might want to look at the timings from your systems to the stratum-2 servers and/or from the stratum-2 servers to the outside world. -- These are my opinions. I hate spam. -- Message: 8 Date: Thu, 4 Oct 2012 18:50:31 +0200 From: Attila Kinali att...@kinali.ch To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] Tracking NTP displacement and correlation between two clients. Message-ID: 20121004185031.b49767822c28ff64c18b5...@kinali.ch Content-Type: text/plain; charset=US-ASCII On Thu, 04 Oct 2012 09:46:38 -0700 Hal Murray hmur...@megapathdsl.net wrote: If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. Within a LAN, RTT is usually in the range of 200us with a jitter in the same range (it can happen that jitter is significanlty higher than RTT, if you have bursty traffic in your LAN) Attila Kinali -- There is no secret ingredient -- Po, Kung Fu Panda
Re: [time-nuts] 5071A Cs Oven
Hi This is one that's been close at hand for the last 15 years or so Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Stan Sent: Thursday, October 04, 2012 1:58 PM To: time-nuts@febo.com Subject: Re: [time-nuts] 5071A Cs Oven If this is the 5071A on that online auction site, I found out from the seller that the error log shows Cs oven timeout. This might be because of a bad oven supply (expensive) or a bad tube (really expensive)! Stan -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of time-nuts-requ...@febo.com Sent: Thursday, October 04, 2012 9:51 AM To: time-nuts@febo.com Subject: time-nuts Digest, Vol 99, Issue 30 Message: 6 Date: Thu, 4 Oct 2012 09:45:22 -0700 From: Tom Van Baak t...@leapsecond.com To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] 5071A Cs Oven Message-ID: CCE1197699D24CBDB6AC7D386B504875@pc52 Content-Type: text/plain; charset=iso-8859-1 Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob -- Message: 7 Date: Thu, 04 Oct 2012 09:46:38 -0700 From: Hal Murray hmur...@megapathdsl.net To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] Tracking NTP displacement and correlation between two clients. Message-ID: 20121004164638.c0fee800...@ip-64-139-1-69.sjc.megapath.net Content-Type: text/plain; charset=us-ascii bow...@gmail.com said: Due to reasons I really can't go into, a systems user is concerned with the displacement of two servers from the same pair of stratum 2 NTP servers. I'm convinced that it really isn an issue as long as the two systems in question remain within a few 10's of ms. However, I have no off the shelf method of collecting and correlating the data. Before I go out and invent the wheel, I thought I would check and see if anyone has done such a thing and saved the scripts and whatnot. Assuming you are running the standard ntpd... It includes all sorts of logging. Set up the two systems so they use each other as servers. Turn on rawstats. ntpd will add a line each time it exchanges a pair of packets with a server. That line will have the IP Address and 4 time stamps. See the documentation. Details are in monopt.html The 4 time stamps are: time the request left the local system time the request arrived at the remote system time the response left the remote system time the response arrived at the local system If you subtract the first two, you get the network transit time for the request packet as skewed by the clock offset. Subtracting the last two gives you the transit time for the response packet. If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. (If you assume the clocks are both accurate, you can compute the network transit time in each direction.) If you want to graph the results, you have to split out the lines for the server you are interested in. Then you can feed it to gnuplot/whatever. You can also do the monitoring from another system, but then you have to sort out what happens when the clock on that system is off. How good is your connection to the big bad internet? If you run a big download over a slow link, the queuing delays can confuse ntp. You might want to look at the timings from your systems to the stratum-2 servers and/or from the stratum-2 servers to the outside world. -- These are my opinions. I hate spam. -- Message: 8 Date: Thu, 4 Oct 2012 18:50:31 +0200 From: Attila Kinali att...@kinali.ch To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: Re: [time-nuts] Tracking NTP displacement and correlation between two clients. Message-ID: 20121004185031.b49767822c28ff64c18b5...@kinali.ch Content-Type: text/plain; charset=US-ASCII On Thu, 04 Oct 2012 09:46:38 -0700 Hal Murray hmur...@megapathdsl.net wrote: If you assume the network transit times are equal, you can compute the clock offset. If you are on a LAN, the transit times will probably be tiny on the scale of 10s of ms. Within a LAN, RTT
Re: [time-nuts] Best counter setting for ADEV?
You're right: it is better to put it down correctly: for a theoretical resolution of 10E-14@100K seconds, start with a 1nS counter that takes 1 second samples for 100K seconds and average those 100K samples. You have your resolution and a noise (an error bar) of 3E-12. If you use a 100pS counter and do the same run, you will end up with a 10E-15 reolution and 3E-13 noise, and so on. It turns out that to have a real 10E-14 measure (@100K seconds), not blurred with noise, you must start with a 10pS counter. On Thu, Oct 4, 2012 at 6:41 PM, Bob Camp li...@rtty.us wrote: Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] Best counter setting for ADEV?
Hi What if I only take two single measurements: One at time = 0, the other at time = 100,000 seconds. No averaging, no signal processing just two measurements. I look at the time difference between the two signals at time = 0, and then again a bit more than a day later. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 2:48 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? You're right: it is better to put it down correctly: for a theoretical resolution of 10E-14@100K seconds, start with a 1nS counter that takes 1 second samples for 100K seconds and average those 100K samples. You have your resolution and a noise (an error bar) of 3E-12. If you use a 100pS counter and do the same run, you will end up with a 10E-15 reolution and 3E-13 noise, and so on. It turns out that to have a real 10E-14 measure (@100K seconds), not blurred with noise, you must start with a 10pS counter. On Thu, Oct 4, 2012 at 6:41 PM, Bob Camp li...@rtty.us wrote: Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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
Re: [time-nuts] 5071A Cs Oven
On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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.
Re: [time-nuts] 5071A Cs Oven
Learn something every day here. How to really extend the old dead C life a bit longer. On Thu, Oct 4, 2012 at 3:04 PM, Richard (Rick) Karlquist rich...@karlquist.com wrote: On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://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-nutshttps://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] 5071A Cs Oven
We really need a group buy of 5071A tubes for a couple hundred each. Thomas Knox Date: Thu, 4 Oct 2012 15:10:11 -0400 From: paulsw...@gmail.com To: time-nuts@febo.com Subject: Re: [time-nuts] 5071A Cs Oven Learn something every day here. How to really extend the old dead C life a bit longer. On Thu, Oct 4, 2012 at 3:04 PM, Richard (Rick) Karlquist rich...@karlquist.com wrote: On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://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-nutshttps://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] 5071A Cs Oven
These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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.
Re: [time-nuts] 5071A Cs Oven
Rick Well I have the old 5060 and 61 tubes and indeed they run out O gas. I am using a 5060 tube in a 5061 and the Frankenstein works though it has only a few Cs. But heck it all locks after 48 hours. So someone said with humor lets buy some 5071 tubes for a couple hundred each. Guess I could have some fun getting that into the 5061. Suspect that would be the bride of Frankenstein then. Thanks for sharing the insights and experience. Paul WB8TSL On Thu, Oct 4, 2012 at 3:21 PM, Richard (Rick) Karlquist rich...@karlquist.com wrote: These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://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-nutshttps://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] Best counter setting for ADEV?
If you take 2 samples out of a 1nS counter than you can estimate your measure with a 500pS resolution (5E-10) with an uncertainty (noise) of 7E-10. On Thu, Oct 4, 2012 at 8:58 PM, Bob Camp li...@rtty.us wrote: Hi What if I only take two single measurements: One at time = 0, the other at time = 100,000 seconds. No averaging, no signal processing just two measurements. I look at the time difference between the two signals at time = 0, and then again a bit more than a day later. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 2:48 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? You're right: it is better to put it down correctly: for a theoretical resolution of 10E-14@100K seconds, start with a 1nS counter that takes 1 second samples for 100K seconds and average those 100K samples. You have your resolution and a noise (an error bar) of 3E-12. If you use a 100pS counter and do the same run, you will end up with a 10E-15 reolution and 3E-13 noise, and so on. It turns out that to have a real 10E-14 measure (@100K seconds), not blurred with noise, you must start with a 10pS counter. On Thu, Oct 4, 2012 at 6:41 PM, Bob Camp li...@rtty.us wrote: Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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
Re: [time-nuts] 5071A Cs Oven
I've read that the 5071A has an almost indefinite tube life. However, the 'high stability' option only has a 9 year life. What happens after 9 years? The tube stops working, or it degrades to the normal tube performance? Tom WB6UZZ Sent from my HTC Inspire™ 4G on ATT - Reply message - From: Richard (Rick) Karlquist rich...@karlquist.com To: Tom Van Baak t...@leapsecond.com, Discussion of precise time and frequency measurement time-nuts@febo.com Subject: [time-nuts] 5071A Cs Oven Date: Thu, Oct 4, 2012 12:21 pm These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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] 5071A Cs Oven
Hi Rick, What I've read or heard is that high-perf 5071 tubes last about 7 years (and then run out of cesium). However, the standard-perf tubes last much longer, maybe 20 years or more (and eventually fail for reasons other than ovens). But there have been different generations of tubes so I'm not sure if these numbers apply, say, to what Bob has. You may be in a much better position than I to confirm any of this. There have been some papers on the subject. One I remember is: http://tycho.usno.navy.mil/ptti/1997/Vol%2029_06.pdf /tvb - Original Message - From: Richard (Rick) Karlquist rich...@karlquist.com To: Tom Van Baak t...@leapsecond.com; Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Thursday, October 04, 2012 12:21 PM Subject: Re: [time-nuts] 5071A Cs Oven These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. 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.
Re: [time-nuts] Best counter setting for ADEV?
Hi If I have two clocks: 1) I know they are X ns apart at time = 0 2) I know they are Y ns apart at time = 100,000 seconds 3) The resolution of the X and Y measurements is 1 ns I know the relative time between the two clocks to a lot better than 5.0x10^10... Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 3:43 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? If you take 2 samples out of a 1nS counter than you can estimate your measure with a 500pS resolution (5E-10) with an uncertainty (noise) of 7E-10. On Thu, Oct 4, 2012 at 8:58 PM, Bob Camp li...@rtty.us wrote: Hi What if I only take two single measurements: One at time = 0, the other at time = 100,000 seconds. No averaging, no signal processing just two measurements. I look at the time difference between the two signals at time = 0, and then again a bit more than a day later. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 2:48 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? You're right: it is better to put it down correctly: for a theoretical resolution of 10E-14@100K seconds, start with a 1nS counter that takes 1 second samples for 100K seconds and average those 100K samples. You have your resolution and a noise (an error bar) of 3E-12. If you use a 100pS counter and do the same run, you will end up with a 10E-15 reolution and 3E-13 noise, and so on. It turns out that to have a real 10E-14 measure (@100K seconds), not blurred with noise, you must start with a 10pS counter. On Thu, Oct 4, 2012 at 6:41 PM, Bob Camp li...@rtty.us wrote: Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin Rosenberg ke...@rosenberg.net wrote: On Oct 3, 2012, at 6:52 PM, Kevin Rosenberg ke...@rosenberg.net wrote: Nice plot! Yes, I'd have trouble measuring 10E-14 at 10E5 seconds. Sorry, 1E-14 at 1E5. ___ 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] 5071A Cs Oven
See also this posting by Dave Carlson (ex hp): http://www.febo.com/pipermail/time-nuts/2005-May/018354.html /tvb ___ 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] 5071A Cs Oven
Rick can correct me on this... but my understanding is that it is almost never a case of running out of Cs from the oven, but that the dynodes of the electron multiplier become covered in used Cs and so the ultimate SNR and effective beam current falls to a point where there is no dependable beam current left to do much with. The tube that I sent to Paul, he was able to get some more current by running the Cs oven at a higher temp and as such is vaporizing off LOTS of Cs and can get a tiny bit of beam current to work with. But it is a down hill slide, since to keep the beam current up with a dirty E-mult, over time you need to keep running up the Cs oven temp to the point where you may just run out of Csor get an E-mult that is so covered with used Cs that there is no beam current to work with. You can always try and run up the E-mult voltage as well, but you then run the risk of arcing in the E-mult and an issue of burning up the resistor divider stack that I believe is part of the E-mult. I can see it now.. an end-of-life Cs tube that needs 35kV on the E-mult to get any beam current to work with! : -) -Brian, WA1ZMS On Oct 4, 2012, at 3:41 PM, paul swed paulsw...@gmail.com wrote: Rick Well I have the old 5060 and 61 tubes and indeed they run out O gas. I am using a 5060 tube in a 5061 and the Frankenstein works though it has only a few Cs. But heck it all locks after 48 hours. So someone said with humor lets buy some 5071 tubes for a couple hundred each. Guess I could have some fun getting that into the 5061. Suspect that would be the bride of Frankenstein then. Thanks for sharing the insights and experience. Paul WB8TSL On Thu, Oct 4, 2012 at 3:21 PM, Richard (Rick) Karlquist rich...@karlquist.com wrote: These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://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-nutshttps://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
[time-nuts] 5071A Cs oven
Hi, Cs depletion in modern tubes is a real failure mode! However some competing modes are: Ion pump failure (whiskers or high impedance shorts) High Cs backround level (Gettering saturation) Oven or ionizer filament failure (More common on older HP units with the AC excitation, (think of those neat light bulbs with the magnet inside where the filament jumps all over the place, less drastic mechanical flexing caused the filaments to eventually fail due to fatigue.) Slow leaks ruining the vacuum Running the Cs with the ovens off as described WILL delay some of these failures, (it's recommended that the ion pump remain energized however) A 5071A tube could be installed into a 5061A (with proper adaptors) but why? Later HP 5061A and 5061B mini tubes with the same diameter as the 5071A tubes can be installed into the 5071A with some caveats and proper adaptors. Corby Woman is 53 But Looks 25 Mom reveals 1 simple wrinkle trick that has angered doctors... http://thirdpartyoffers.juno.com/TGL3141/506df1798baa27179062fst04duc ___ 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] jackson labs.
Thank you very much, Giovanni. I'm forwarding this informal quote to a group, trust this is OK to do. Very reasonable! Thanks again Don GIOVANNI D'ANDREA Hi Don, Sorry about the delay, 1. PN: 1009302 LC_XO GPSDO W/TCXO $355.00 If you need a formal quote please send me you contact info. Let me know if you need anything else. Thanks, Giovanni -- Neither the voice of authority nor the weight of reason and argument are as significant as experiment, for thence comes quiet to the mind. De Erroribus Medicorum, R. Bacon, 13th century. If you don't know what it is, don't poke it. Ghost in the Shell Dr. Don Latham AJ7LL Six Mile Systems LLP 17850 Six Mile Road POB 134 Huson, MT, 59846 VOX 406-626-4304 www.lightningforensics.com www.sixmilesystems.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.
Re: [time-nuts] 5071A Cs Oven
Brian Funny you mention the dirty emult. At the time I was wondering and if there was any insane way to get the Cs off the e multiplier. HV to ground or something to migrate them. Hold upside down and shake hard. ;-) Regards Paul On Thu, Oct 4, 2012 at 4:22 PM, Brian, WA1ZMS wa1...@att.net wrote: Rick can correct me on this... but my understanding is that it is almost never a case of running out of Cs from the oven, but that the dynodes of the electron multiplier become covered in used Cs and so the ultimate SNR and effective beam current falls to a point where there is no dependable beam current left to do much with. The tube that I sent to Paul, he was able to get some more current by running the Cs oven at a higher temp and as such is vaporizing off LOTS of Cs and can get a tiny bit of beam current to work with. But it is a down hill slide, since to keep the beam current up with a dirty E-mult, over time you need to keep running up the Cs oven temp to the point where you may just run out of Csor get an E-mult that is so covered with used Cs that there is no beam current to work with. You can always try and run up the E-mult voltage as well, but you then run the risk of arcing in the E-mult and an issue of burning up the resistor divider stack that I believe is part of the E-mult. I can see it now.. an end-of-life Cs tube that needs 35kV on the E-mult to get any beam current to work with! : -) -Brian, WA1ZMS On Oct 4, 2012, at 3:41 PM, paul swed paulsw...@gmail.com wrote: Rick Well I have the old 5060 and 61 tubes and indeed they run out O gas. I am using a 5060 tube in a 5061 and the Frankenstein works though it has only a few Cs. But heck it all locks after 48 hours. So someone said with humor lets buy some 5071 tubes for a couple hundred each. Guess I could have some fun getting that into the 5061. Suspect that would be the bride of Frankenstein then. Thanks for sharing the insights and experience. Paul WB8TSL On Thu, Oct 4, 2012 at 3:21 PM, Richard (Rick) Karlquist rich...@karlquist.com wrote: These ideas are interesting. AFAIK, there is very little chance of running out of Cs in many years, long past the time when something else in the tube will have reached its end of life. Where did this idea come from? Certainly, not anyone working at HP, Agilent or Symmetricom. A LONG time ago, if you bought a high performance 5061, the increased Cs beam flow might have been an issue, but the amount of Cs shipped with the tube was increased before the 5071A such that running out of Cs is no longer an issue. If you turn off the Cs, you simply have a 10811 to connect to your GPS. Except this 10811 has 10X the EFC range vs a stock 10811. Again, this doesn't make a whole lot of sense, IMHO. BTW, another interesting idea is degaussing the CBT. According to Len Cutler, this is completely unnecessary in the 5071A. As anyone who knew Len can attest, if he thought something was good enough and didn't need to be improved, you could be sure about that. For example, Len wanted us to use tantalum filter capacitors for energy storage in the power supply. Fortunately, we talked him out of that (actually the $100+ cost helped to convince him). Due to customer demand (the customer is always right :-) the Santa Clara Division eventually released a degausser. One of the reasons why this is not necessary in the 5071A is the Zeeman line measurement of the magnetic C field. The 5061 was open loop. So again there may have been a grain of truth behind the folklore. Rick Karlquist N6RK Member of the 5071A design team On 10/4/2012 9:45 AM, Tom Van Baak wrote: Bob, Right, not good. There should be a fault message associated with it. Check the internal log. OTOH, to conserve cesium, I've heard that some people run 5071A in standby mode most of the time and only turn on the cesium for a fraction of an hour a day or a week (to recal the quartz). Running in this standby mode is also advisable if one wants the best short-term performance out of a 5071A, without the loop noise. In this respect it's just like a GPSDO - you get the best short-term performance if you turn off disciplining. /tvb Hi Is Cs oven: 0.0V a good thing on a 5071A? I suspect not, since either it or GPS is off frequency. Bob __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nuts 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 https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow
[time-nuts] info on Vectron Crystal Oscillator Model 224-5136, 48MHz
I could not find any info on the Vectron Crystal Oscillator Model 224-5136, 48MHz so I decided to open one up and share what I have been able to figure out. Currently I have two questions: supply voltage, and Adj voltage range? There are seven solder lugs in a semi-circle, it looks like a mirror image of the letter C. Counting clockwise, the First lug is 11 O'clock, Second lug 12 O'clock, Third lug 1 O'clock, the Seventh lug is at 7 O'clock. 1 - NC 2 - GND 3 - GND 4 - VCC (positive supply) 5 - NC 6 - Adj 7 - GND On the inside, you can see that one wire (Orange) goes to lug #4. In addition there are two zener diodes voltage regulator running off of #4. There is a 6.2V zener (1n753A) and a 15V zener (1n4109). A Black wire going to #2,#3,#7 and the case. The #6 lug goes to a resistor divider 24K Ohm 100K Ohm to GND, a gray wire from the center of the divider, this is what I am assuming is the fine adjust. Ok, so now it is clear that I need more than 15V supply, but how much? I'm assuming that the orange wire is directly to the Heater, while the 15V and 6.2V would be for the Oscillator and buffer or control circuitry. So I raised VCC in one volt increments and measured across the zeners until they were at the correct voltage. However it is kind of like having two watches at the same time - neither of the zener are at the correct voltage at the same time. It would probably be safe to run at ~20V, but I'm still curious what the heater circuit was designed for. Supply Z6.2Z15 14V5.316 11.73 15V5.566 12.59 16V5.801 13.45 17V5.966 14.31 18V6.135 14.91 19V6.150 15.11 20V6.166 15.33 21V6.175 15.47 22V6.180 15.64 23V6.199 15.83 24V6.210 16.00 --- 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.
Re: [time-nuts] Best counter setting for ADEV?
OK, because you think: (Y-X)/100K is the drift of the two clocks for each second over 100K seconds and for each second I have a 1nS/100K resolution. Now think about this: repeat the measure over and over and after each 100K seconds you have your table. This table (your counter is 1nS) can have figures like, say, 52nS, 53nS, 58nS, that is with a 1nS step. Your drift maybe not exactly 52: it maybe 52.1 52.3 52.4 but you only see 52. So you are loosing an entire nS between 52nS and 53nS. If you have a 100pS counter you can see 52.0, 52.1 and so on. On Thu, Oct 4, 2012 at 10:13 PM, Bob Camp li...@rtty.us wrote: Hi If I have two clocks: 1) I know they are X ns apart at time = 0 2) I know they are Y ns apart at time = 100,000 seconds 3) The resolution of the X and Y measurements is 1 ns I know the relative time between the two clocks to a lot better than 5.0x10^10... Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 3:43 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? If you take 2 samples out of a 1nS counter than you can estimate your measure with a 500pS resolution (5E-10) with an uncertainty (noise) of 7E-10. On Thu, Oct 4, 2012 at 8:58 PM, Bob Camp li...@rtty.us wrote: Hi What if I only take two single measurements: One at time = 0, the other at time = 100,000 seconds. No averaging, no signal processing just two measurements. I look at the time difference between the two signals at time = 0, and then again a bit more than a day later. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 2:48 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? You're right: it is better to put it down correctly: for a theoretical resolution of 10E-14@100K seconds, start with a 1nS counter that takes 1 second samples for 100K seconds and average those 100K samples. You have your resolution and a noise (an error bar) of 3E-12. If you use a 100pS counter and do the same run, you will end up with a 10E-15 reolution and 3E-13 noise, and so on. It turns out that to have a real 10E-14 measure (@100K seconds), not blurred with noise, you must start with a 10pS counter. On Thu, Oct 4, 2012 at 6:41 PM, Bob Camp li...@rtty.us wrote: Hi Maybe we're talking about two different things here. Simplistically, resolution is simply what the counter puts out as a usable LSB. There are a lot of examples out there that will do a one nanosecond single shot measurement. That measurement includes the normal trigger noise and stuff in the counter. There are a few assumptions about slew rate of the input. If I take two one shot measurements spaced 100,000 seconds apart, my resolution over that period is 1.0x10^-14. The measurement is representative of what the sources have done to that level. Weather the sources are stable to that level is independent of the resolution of the measurement. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 12:27 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Yes, for a theoretical resolution of 10E-14@100K sec start with a 1nS and average for 100K seconds but ending up with a noise (an error bar) of 3E-12. On Thu, Oct 4, 2012 at 4:25 PM, Bob Camp li...@rtty.us wrote: Hi For a *resolution* of 1.0 x 10^-14 at 1.0 x 10^5 seconds I only need a 1.0 x 10^-9 second reading out of the counter. Indeed, 5 or 10X more than that would be better if I was after a 1 x 10^-14 accuracy. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Azelio Boriani Sent: Thursday, October 04, 2012 7:58 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] Best counter setting for ADEV? Taking 10 samples from a 1nS counter leads you to a 100pS resolution but the noise at best (if it is really random) is reduced by SQRT(10). So, as already pointed out here, there is no real substitute for lower noise, higher resolution measurements. On Thu, Oct 4, 2012 at 3:01 AM, Bob Camp li...@rtty.us wrote: Hi A 1 ns resolution TI counter will do the measuring part just fine. Hitting the number, is where it gets a bit crazy. A *good* GPSDO might get near that. Bob On Oct 3, 2012, at 8:56 PM, Kevin
