[time-nuts] 10 MHz reference switching
Hello Time Nuts, I'm back again. I designed bandwidth-limited (30 MHz) ADA4899-1 opamp circuits to buffer the 10 MHz reference, with theoretical noise of less than 3 nV/rtHz. Oscillator noise is about -155 dBc/rtHz from 1 KHz to 10 KHz, and -160 dBc/rtHz from 10 KHz to 100 KHz. Unfortunately the rest of the design is holding up PCB manufacture, so results on these won't be available for awhile. Which brings me to the other design-stopping considerations: Is anybody aware of low-noise analog switches that can be used to reconfigure the 10 MHz reference paths without significantly degrading the phase noise? As always, I believe 1/F noise would be the concern. Non-reflective is great, but I can accomplish that by using more switches and 50 ohm resistors if need be. Isolation is a key consideration as well. Mechanical relays are out of the question. An alternative to switching is phase-locking the OCXO to an external reference. I have thoroughly looked into this possibility, and it is likely my preferred approach. But I need to understand how the switching approach compares, hence my question. I do not need help with phase-locking; this feature is readily available from OCXO vendors who have dedicated years of research to optimizing their products. Any switch experience out 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] 10 MHz reference switching (Clay)
> Why would flicker noise be a concern if there is no dc current flowing > in the switches? > Bruce Hi Bruce, Does it matter if the current is DC or AC? I expect to see +/-20 mA current due to the signal level. Clay From: Bob Camp Subject: Re: [time-nuts] 10 MHz reference switching > Hi > Or put another way: > If you have a structure that uses series resistors and shunt diodes: > The diodes are reverse biased when the switch is "on". > The diodes are forward biassed when the switch is "off". > No current when on = no noise. > Bob Hi Bob, I had not considered pin diode switches. I guess in my mind they are "RF", and 10 MHz is not. No reason not to think about them. However, I think to get any sort of useable isolation at 10 MHz series-shunt or pi configuration is required, which will have current flowing in one of the paths at all times. Clay From: paul swed Subject: Re: [time-nuts] 10 MHz reference switching > rf relay would work Hi Paul, Sorry, no relays for a high-reliability application. Not enough room or current either. Clay From: Bob Camp Subject: Re: [time-nuts] 10 MHz reference switching > Hi > Assuming you are switching between an external reference and the internal > OCXO: > The "bleed through" of the OCXO is going to show up as a discrete spur close > to the external reference carrier. Unless you have a very > unusual > architecture, it will be inside the loop for everything you are doing. > Attenuating it to a "suitable" level may be quite difficult. > Bob Good point. I could power down the internal oscillator based on a detected external reference. Just more circuits to add . . . Thanks to all who replied, you raised some good points. Locking the internal to external sounds better all the time, it's just kind of expensive. Clay ___ 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] Realistic Tbolt phase noise
> From: "Bob Camp" > Subject: Re: [time-nuts] Realistic Tbolt phase noise > Hi > I think you will find that the noise floor of the ADF4107 (dividers and > phase detector) is a bit higher than -150 at 100 Hz offset at 10 MHz. It > should get you past the -120 level though. That's still much better than you > are seeing. > Indeed -120 is not what figure 15 on page 8 of the specification would > imply. I'm sure that the figure is technically correct. It's just not > showing the complete picture. > Bob I have found the Analog Devices chips to have a very poor phase comparator noise floor close-in. If you do the math based on their published figure of merit, it won't add up. I believe the issue is 1/F noise due to the low-current CMOS process they use. They are cheap, jelly-bean chips that have their uses. Definitely not high performance. Bipolar is they way to go. Clay ___ 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] GPS 1 PPS accuracy limits
Time Nuts; I have a customer request for a microwave frequency synthesizer with extreme accuracy requirements; 4 X 10^-11, or 0.04 PPB. Obviously this is way out of quartz oscillator territory. Is GPS 1 pulse-per-second useable, or do they need an atomic clock? Maybe they don't realize what they are spec'ing. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS 1 PPS accuracy limits
Date: Tue, 20 Apr 2010 18:59:20 -0700 From: jimlux Subject: Re: [time-nuts] GPS 1 PPS accuracy limits >Do they have other requirements (Allan deviation? phase noise?) >Over what time span do they want 4E-11 accuracy? 1 second, 1000 seconds, >days? >A synthesizer locked to a Stanford Research PRS10 (datasheet accuracy >5E-11) is in that ballpark. They have phase noise requirements, which aren't achievable for the price they want to pay. But phase noise is determined by the OCXO and other circuitry. I suspect the most likely answer is they wrote the spec wrong, and don't really mean 1 Hz absolute accuracy. I'll talk to them and clarify. ___ 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] Advice on 10 MHz isolation/distribution amplifier
Hello everyone, I am new to this list, happened across it while searching on distribution amplifiers. I need to design a 10 MHz isolation/distribution amplifier with two outputs for a high-vibration wide temperature range environment. I was considering using a design based on the NIST article 'A new 5 and 10 MHz distribution amp', most likely the third incarnation, as I cannot tolerate modulation of the signal via the transformer. I have looked at the work by Gerhard Hoffman, which is impressive. However, the transformer remains. I should mention I am also limited to 15V power supply, and need to output into 50 ohms at 10 dBm. Ideally DC current will be minimized, preferably 20 - 30 mA per channel. I can generate a negative power supply if necessary, but would prefer not to. If anybody can offer tips so I don't undertake to reinvent the wheel, it would be appreciated. Thanks in advance, - Lifespeed ___ 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] Advice on 10 MHz isolation/distribution amplifier
Avoiding transformers and inductors will make it virtually impossible to achieve very low phase noise as the dc gain from say the base of any transistor in the chain to the output will degrade the flicker phase noise. Using transformers or using an inductor to shunt any collector resistors reduces the flicker phase modulation to low levels. JPL in the past has built capacitively coupled complementary symmetry isolation amplifiers that avoid transformers but suffer from dc loop gains of around 3 or so. Using complementary symmetry can be a good way of keeping the dc current down. How much reverse isolation do you need? How low does the phase noise floor need to be? What about flicker phase noise, how low does that need to be? Bruce Right, what do I really need? I only have a really good 10 MHz OCXO crystal oscillator to distribute, so about -120 dBc at 10 Hz, -140 dBc/Hz at 100 Hz, - 150 dBc/Hz at 1KHz, and -155 dBc/Hz noise floor. No maser or cesium clock, living in the world of practical realities here. Of course I would like to be 3 - 6 dB better than the OCXO numbers. Reverse isolation is my primary interest in the distribution amplifier approach, although the OCXO is good enough that a sloppy approach could contaminate the phase noise also. I would like to accomplish at least 100 dB reverse isolation at frequencies below 20 MHz, but more is better in this case. The 10 MHz is running all over a noisy aircraft, to potentially noisy receivers. In reading up on the subject, I have come to understand that DC gain is the bane of close-in phase noise. Given that flicker noise is such a headache for we frequency synthesizer designers, I guess this should come as no surprise. Clay (AKA Lifespeed) ___ 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] Advice on 10 MHz isolation/distribution amplifier (Clay)
Message: 2 Date: Thu, 11 Feb 2010 07:54:40 -0500 From: Bob Camp Hi Implementing that circuit without using a hybrid would be a bit of a challenge. Bob Message: 6 Date: Fri, 12 Feb 2010 04:09:08 +1300 From: Bruce Griffiths Yes implementing an exact copy without using a hybrid would be difficult. However for 10MHz use, its probably not too difficult since that isolation amplifier is intended for a 100MHz signal and the requirement is for 10MHz operation. If the transistor ft's are reduced by a factor of 10 or so it shouldn't be too much of a problem. At 10MHz 2N3906 and 2N3904 transistors should suffice. Bruce Hi Bruce, Thanks for the tips. I've been trying to follow the circuits you posted. The first one, in .PNG format, looks like a common-base complementary (push-pull) stage followed by a common-emitter complementary stage to provide the low impedance output. The second circut in .GIF fromat I am having a bit more trouble understanding. I see that V6, 7 are at the outputs and just used for to simulate isolation. V1 is the input? Are Q5,6 used to set the bias point of Q4? Are V2,3,4 just there to bias the transistors for simulation purposes, and this would be accomplished another way in a real implementation? Please explain the comment regarding the hybrid. Are you and Bob referring to a 90 degree hybrid coupler, or other quadrature method like a transmission line transformer? What would be the purpose of such a device? Would it be too much to ask for a description of these circuits? I suppose we all have our areas of expertise, and transistor isolation amps are somewhat new to me. Thanks again for all the help. Clay PS - yes, the OCXO is vibe isolated. And you are certainly correct about long runs of single-ended coax being susceptible to noise. The system designer has accepted this and allowed for some degradation. But I will look into the practicality of implementing a differential line for the long run of 10 MHz cable. However, I will still need to implement traditional coaxial isolated 10 MHz outputs. ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 6 Date: Thu, 11 Feb 2010 12:38:58 -0500 From: "Bob Camp" Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution amplifier(Clay) Hi I'll grab the one on the hybrid. In this case hybrid is referring to a construction technique. The circuit shown was originally fabricated in a TO-8 package with chip and wire construction. It was certainly made using thin film or thick film technology on a substrate. Based on the number of components and size of the part, I'd bet that the resistors were printed on the substrate. When you are using that kind of construction approach there are some good things that happen and some bad things. The circuit topology is modified to work with the construction technique. In this case the Ft's of the transistors are quite high. Taming them on a substrate (alumina or glass) is a very different thing than doing it on a PC board. --- Is your OCXO vibration isolated? -- Bob OK, you're talking about chip-and-wire, or hybrid construction. Really, at 10 MHz, that seems unnecessary. Of course, PCB construction will introduce some parasitics. Transistors with lower Ft could be used. Additionally, there are bandwidth-limiting techniques like adding feedback capacitance. I suppose this would come at the expense of high-frequency isolation. But for my application isolation is important in the 10's of MHz, not much higher. Having built a couple transistor circuits over the years, I am aware of what can happen when a transistor in a low frequency circuit has an Ft of multiple GHz. Usually oscillations. I imagine a high Ft enables the isolation to extend to very high frequencies. Yes the OCXO is vibration isolated or the system wouldn't produce good phase noise in operation. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 1 Date: Thu, 11 Feb 2010 12:42:27 -0500 From: "Bob Camp" Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution amplifier(Clay) Hi I really should learn how to read the whole message Cancel the second request on vibe info. - The gotcha with vibration isolation is that it will stop working at some lower frequency. Aircraft have plenty of vibration running around at low frequencies. That all sounds like bad news. Actually it's not. Since the phase noise isn't going to be all that good below the cutoff of the isolation, the amp doesn't need to sweat super low phase noise very close in. That can make the choice of transistors easier. Bob Thanks Bob. I am aware of all the vibe issues, low freq corner, resonance peaking, etc. And yes, I have seen the Wenzel spreadsheet. Wenzel is a good resource for info. These issues have all been looked into. The phase noise numbers are what is predicted under vibration (10 Hz number might degrade a few dB). The amplifier will need to be better. 1 Hz < -100 dBc/Hz 10 Hz < -125 dBc/Hz 100 Hz < -140 dBc/Hz 1 KHz < -150 dBc/Hz 10 KHz < -155 dBc/Hz Are you aware of any bipolars that are better than others in 1/F noise performance? I noticed Gerhard Hoffman's design used BFG198 and BFG31, although those are SOT223 parts, which are somewhat large for my design. If I'm not mistake 'low saturation' correlates to low 1/F noise . . . I simulated the circuit with two outputs you sent in .GIF format. It appears to be tuned to a somewhat lower frequency than 10 MHz, perhaps 10 KHz to 1 MHz where the overall gain is near 0 dB, and the phase shift is near 0. I am using MMBT3904 transistors with Ft near 250 MHz. Perhaps that is the issue. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 2 Date: Fri, 12 Feb 2010 12:12:29 +1300 From: Bruce Griffiths The output (collectors of Q5, Q6 emitter of Q4) of the input amplifier sets the dc voltage at the inputs ( Q1 base, Q7 base respectively) of the output amplifiers. The circuit consists of a unity gain input amplifier (Q4, Q5, Q6) that drives a pair of output amplifiers (Q1, Q2, Q3 and Q7, Q8, Q9 respectively) each with a gain of 2x (6dB). The input amplifier is essentially a white emitter follower with a complementary symmetry output stage (shown in transistor electronics books from the 1960's) where an input CE transistor drives a complementary pair of CE transistors with feedback from the common collectors of the 2 output transistors to the input transistor emitter. In effect its merely a very simple unity gain opamp. Its usually best to ensure that the CE output stage pair provide the dominant open loop pole. Using a higher ft (2 to 3x) input transistor than the output pair is the usual way of ensuring this. Well, it is so obvious now that you explained it. I had forgot about the need for one of the stages to set the dominant pole. Thanks Bruce and Bob for sharing your obsession with frequency controls. I'll simulate this further, and have a prototype PCB built within the next few weeks. I did notice the resistor at the base of Q2,5,8 is responsible for significant noise. I'll have to be careful with the bias circuit. Have to get busy for now, but I will report back with results. Best regards, Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 3 Date: Thu, 11 Feb 2010 23:02:30 -0500 From: Bob Camp Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution (Clay) Hi I suspect your noise spike can be cured by a series R-C to ground from the junction of Q1 base, Q7 base and all the other stuff. Something is going to have to set a high frequency roll off. With no coils some combo of R and C is going to have to do it. You might also try returning all of the upper emitter resistor bypasses to ground rather than B+. Another alternative would be emitter to emitter bypass as shown on the JPL schematic. I'm guessing both would improve isolation in a real world circuit. Bob Hi Bob, Please provide a link to the JPL schematic. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
I seem to not be accomplishing much isolation from output to input, as well as output to output. Have I fumbled PSPICE somehow? For each simulation, Vac was set separately, with V1=0.707V at the input, while V6=0V at the output (sim1). Then V1=0V, and V6=0.01V (sim2). clay_schem1.pdf Description: Adobe PDF document clay_sim1.pdf Description: Adobe PDF document clay_sim2.pdf Description: Adobe PDF document ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 6 Date: Thu, 18 Feb 2010 06:53:47 -0800 (PST) Subject: [time-nuts] Advice on 10 MHz isolation/distribution SRS FS700 An easy way to make a 10 MHz distribution amplifier is the one from Stanford Research. Manual can be find at : http://www.thinksrs.com/downloads/man.htm Download FS700 The schematics are in the back of the manual. Met vriendelijke groeten Regards Thank you, but I require a circuit without transformers for reasons of vibration susceptibility. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Date: Thu, 18 Feb 2010 15:22:40 +1300 From: Bruce Griffiths Clay What's the effect of assigning the same label (Vout_2) to the outputs of both output amplifiers as shown in your circuit schematic? Bruce Hi Bruce, Duplication of the Vout_2 net labels (copy and paste is great) did not help the isolation. Neither did the 10 ohm base resistors. Now that the schematic is correct, the base resistors do not ruin the noise performance. Sigh . . . So this circuit looks like it will work well (I also get 100 dB simulated isolation between outputs, as well as noise of less than 2 nV/rtHz). I do believe isolation is important, as there are several subsystems using (and possibly trying to contaminate) the reference. I don't know all the details of each subsystem, and may not be privy to such info. So I can't put a number on the required isolation, but 20 log N applied to spurious within the PLL bandwidth is very unforgiving. The opamp solution is appealing, but I am unsure if it will work. My favorite low-noise parts right now are the TI OPA211, and the Analog AD797. I know the AD797 can be a bit tricky to tame oscillations. They both are probably difficult when configured for more than 10 MHz BW. I will run some simulations, but at 2V p-p into 50 ohms (10 dBm) the required rails would need to be more than 5V, although I have 15V available. I am unsure of the distortion performance at 20 mA peak output drive. Although the AD797 can source more current, neither amp is specified with good distortion numbers at 50 ohm loads. They always show audio (600 ohm) loads for their good THD numbers. I need to provide harmonics lower than 30 dBc. At first glance it does not look a likely solution. Lastly, my customer, the system designer, would now like to be able to add a switch function to the 10 MHz distribution. I will have to check and see if switching of bias current to these two-stage transistor amp circuits can accomplish this function. Jelly-bean analog switches are unlikely, I think, to have the requisite noise and isolation performance. Nothing like hitting a moving target. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 5 Date: Thu, 18 Feb 2010 11:11:18 -0800 From: Hal Murray Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution (Clay) > Thank you, but I require a circuit without transformers > for reasons of vibration susceptibility. Is the problem wires moving relative to eachother within the transformer or the whole transformer moving relative to the local magnetic field? Does it help to pot things in epoxy or varnish or ...? Years ago, when I took the lid off a small mini-circuits transformer it was just a few turns on a toroid. Something like varnish would be easy to apply. Hal, The problem is modulation of the reference signal via relative movement of the transformer guts. While I understand there are ways to 'harden' magnetic devices, my application is far too sensitive to even consider a magnetic approach given the availability of alternatives. Any spurious signals present on the 10 MHz reference will be increased by 56 dB by the phase-locked loop. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 5 Date: Thu, 18 Feb 2010 17:13:55 -0500 From: "Bob Camp" Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution (Clay) To: "'Discussion of precise time and frequency measurement'" Message-ID: <13d264786a8a41e781bb4e53c1dd9...@vectron.com> Content-Type: text/plain; charset="iso-8859-1" >Hi > > Have you looked into any actual data on small rf transformers under > vibration at 10 MHz? Hi, No. I have experience with low-noise designs under vibration, however. What is the compelling reason for even considering such a device when a design is available that doesn't need it? Sounds like risk without an upside to me. >The phase noise of your reference is going to head up by your +56 db. As >long as the transformer's impact is well below the reference, it will not >show up in the output signal. any spur larger than -130 dBc will be an issue. It doesn't take much vibe to do that. And the synthesizer has to operate under 10g RMS. >Noise at the load is a problem, no matter how it gets there. Transmission >may not be inside your scope -yet. It's still an issue that directly impacts >what you are trying to do. Only one device is at the end of a long cable, with reduced performance expectations. The others are local and share a common ground. I am familiar with the various types of capacitors properties and frequencies. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
The TI THS3201-EP was looking pretty good for a high speed opamp. But the input current noise graph doesn't go below 100 KHz and is climbing pretty steeply at that point. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
Message: 6 Date: Fri, 19 Feb 2010 15:17:44 - From: "Garry Thorp" Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution Hello Clay, Joining in this discussion at a rather late stage - have you considered using 74AC series gates as buffers? They provide reasonable isolation and have surprisingly low phase noise. Hi Gary, That's funny! I have already done everything you describe, and shipped synthesizers with such circuits in them. I like Fairchild UHS inverters for the logic gates. And yes, I've done the low-noise regulators constructed from voltage references, heavy filtering, and an opamp driving a BJT pass transistor (I like the NZT560A/660A). I will mention that for some applications, the Burr-Brown Reg104A is one of the few IC regulators that has suitably low noise for frequency synthesis. However, at the very limits of phase noise performance, the squaring technique is inferior to a sine wave. Not by alot, but enough that I need to stick with a sinusoid. Not to mention the other modules in the system are expecting a sine wave. Thanks, Clay ___ 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] Advice on 10 MHz isolation/distribution
Any opinions on the suitability of the TI OPA820 as a 10 MHz distribution amp? The voltage noise is only 7 nV/rtHz @ 100 Hz (gain=2, I'll be using g=1), which is the best I've seen so far for a wideband amp. Not as good as the discrete transistor circuit, but it would use alot fewer parts. It only needs to be better than a 'good' OCXO. Clay ___ 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] Advice on 10 MHz isolation/distribution
Message: 6 From: Pete Rawson Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution >Clay, > >Analog devices AD4899-1 voltage noise = 2nV/rtHz @ 10Hz; GBW = 300MHz. > >Pete Rawson That is an interesting part. So, how does one think about reverse isolation in a feedback amp? Is the open-loop gain of the amp applied to the spurious signal to cancel it? The spurious, reduced by the ratio of the 50 ohm source impedance, to the output impedance of 0.7 ohms of the amp (-18 dB) appears at the inverting input . . . And what would appear at the non-inverting input? Off to the simulator. Clay ___ 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] Advice on 10 MHz isolation/distribution
If I can believe the simulation (a big if) the ADA4899-1 can provide 90 dB isolation at 10 MHz, rolling up to 70 dB at 100 MHz, when configured with a gain of +2 The subcircuit model provided by ADI is useless for noise, unfortunately. I guess I'll have to build and measure, which I don't mind doing. Unless it turns out to be a waste of time, and I should have just used the transistor circuit. Clay ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Phase Noise of 74AC gates
>From: "Martyn Smith" >Subject: [time-nuts] Phase Noise of 74AC gates > >I'm not about to say how we achieve -170 dBc/Hz phase noise, but we >definitely don't use op amps or logic gates!! > >Regards > >Martyn Is this an advertisement? I don't think anybody believes a logic gate provides state-of-the-art reference buffering. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] OT: HP 8590A
Date: Fri, 19 Feb 2010 17:38:50 -0700 Subject: [time-nuts] OT: HP 8590A Since the list members are familiar with lots of test equipment, I'd like to ask what the folks here think about the HP 8590A Spectrum Analyzer. Is this model ok? Are there any particular failures I should be aware of in this 20+ year old equipment? I have a chance to buy one locally. The only option is has is GPIB. I took a preliminary look at it and it passes the simple test/cal procedure from chapter 1 of the Ops manual. This model only goes to 1.5GHz, but would still be useful for Amateur use. I do wish it would go up to 3GHz, however. I have never owned a spec an, but am somewhat familiar with their usage. Thanks for the input. Joe KA5ZEC I personally do not like these low-end spectrum analyzers. They have poor dynamic range and phase noise performance. However, I design microwave circuits for a living and can be a test equipment snob. If you think it is adequate for your purposes, I would at least connect it to a calibrated signal generator and verify amplitude accuracy is within 3 dB. Most old spec ans I have seen are way off, even broken. Still show a signal on the display, but not very helpful. Also check for spurious across many frequencies. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
From: Bruce Griffiths Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution Its easy to calculate a lower bound to the amplifier phase noise floor from the signal level at the output and the amplifier output noise due to feedback resistors together with the amplifier input voltage and current noise. The actual phase noise floor may be a few dB higher. Bruce Yes, it is easy to calculate. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
From: Pete Rawson Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution To: Discussion of precise time and frequency measurement Message-ID: <1027d446-fe23-45a7-88ad-1cfd339c4...@earthlink.net> Content-Type: text/plain; charset=us-ascii Clay, I would really like to have a copy of your test board for this effort. I'm sure it will be a challenging layout. But it could be a very handy piece of hardware. Supporting a PCA order would easy to do. Pete Rawson I always buy more than one PCB, 5 - 6 don't cost any more than 1. Unfortunately, I don't think my favorite prototype shop (PCB123) will be suitable for a layout with voids in the ground plane (reduce parasitic C) under the opamp and R4003C on the top layer, so I will have to do a 'real' PCB. I was also planning on using the leadless version of the part, which could be challenging for home assembly. Everything is leadless these days. Gives me fits sometimes. I just went through a nightmarish experience with a leadless ARM7 processor that needed some wires added. I replaced it with a leaded part at a cost of 2 mm PCB area . . . but I digress. Clay ___ 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] Advice on 10 MHz isolation/distribution (Clay)
From: Bruce Griffiths Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution (Clay) The calculation is useful to help eliminate amplifiers that will not meet the phase noise floor specs. The amplifiers that remain will then need to be tested/measured for their actual phase noise. Bruce I have been thinking about how to make the measurement. I don't normally measure OCXO phase noise. I buy them to spec and lock my synthesizer to them, measuring phase noise at microwave frequencies using an older Agilent E5500 system with an 8254 (?) signal generator reference, which is an older version of their 8257. I was thinking I could split an OCXO signal, run one path thru the amp, and then mix them together and put the IF output into the E5500. I think I would need a phase shifter to ensure the signals are in quadrature at the mixer. I guess if I had two OCXOs I could let the E5500 control the EFC port on one, it would then use it's internal PLL to lock the 'reference' OCXO to the 'DUT' OCXO/amp. ___ 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] Are the days of buying a crystal numbered?
From: Henry Hallam Subject: Re: [time-nuts] Are the days of buying a crystal numbered? Anybody know what its performance is like under vibration? Henry Just guessing here . . . poor Unless it is a specifically G-hardened device, it is sure to be microphonic. All crystals are. G-hardened is just somewhat better. Clay ___ 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] Advice on 10 MHz isolation/distribution /
From: "Garry Thorp" Subject: Re: [time-nuts] Advice on 10 MHz isolation/distribution / Phase Noise of 74AC gates The attached plot shows the 12.5MHz phase noise plus that of the 100MHz OCXO. The divider phase noise tracks 18dB below the OCXO at low offsets as expected, before its flicker noise and eventual noise floor predominate. I was primarily interested in seeing what the flicker noise was like, but I was surprised when I saw how low the floor was! The E5052B does the necessary calibration automatically before doing a measurement - from my experience with the instrument I have no reason to doubt the validity of the result. (The indicated 100MHz phase noise in the ~1-50kHz region is actually limited by the E5052B, owing to my setting only 100 correlations. However it shows it low enough to indicate that the CMOS noise dominates over that range.) I realise that a divider is very different from a simple inverter, but I think this gives an useful indication of what AC logic is capable of. Garry Pascall Electronics Ltd - Registered in England No: 1316674 VAT Registration No: GB 448705134 Registered Office: Brunswick Road, Cobbs Wood, Ashford, Kent, TN23 1EH Gary, Hi Gary, Interesting plot, thanks for sharing. Like I said previously; logic gates absolutely have their uses in reference buffering. I have used them before, and will continue to do so. However, I think in the most demanding applications the relatively-poor 1/F noise of CMOS loses out to bipolar transistors. If one wanted to stick with a square wave reference, or implement a digital phase detector or divider other types of logic like ECL are more appropriate. Also, digital logic, even high-current ECL, is more susceptible to AM to PM conversion from power supply noise. It all depends on what you need to do. CMOS is pretty good, but can't get you that last bit of performance especially close in. Which is, after all, what a clean reference is for. Clay ___ 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] help me understand AM noise
Hi Guys. It has been a while since I posted, hope you can help with a slightly time-related topic. Can't have frequency without amplitude . . . I recently designed an Automatic Level Control circuit consisting of dual-slope detector logger, open and closed loop references with AM modulation, and a linearizer (volts/dB) driver for series/shunt microwave attenuators. This is part of a DC - 20 GHz microwave synthesizer. I measured the AM noise at 3 GHz, both open and closed loop, and find the noise level is higher at the output of the attenuator/amplifier chain at similar power levels to the input (13 dBm). The input RF chain saturates at about 17 dBm, while the output amp following the attenuators saturates at about 20 dBm. I understand that an amplifier in compression will suppress AM noise. What I wonder is are my measurements of increased AM noise (red trace) at the output of the attenuator/amp lineup to be expected based on the higher available saturated power? Is it possible to attenuate the signal using the power control (open loop in this example, ALC is not used) without degrading AM noise performance? Does anybody have any suggested reading on this subject? I am trying to understand how well my circuit performs, in general. I do observe that control the power to a lower level increases the AM noise. But it is a relative measurement to begin with, so what is "good"? I have been reading the Agilent E5500 user guide on AM noise measurements, but don't find a great deal of information there regarding AM noise performance of a Device Under Test. Thanks, Lifespeed http://home.comcast.net/~claybu/pics/electronics/am_noise_1.png ___ 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] help me understand AM noise
Thanks for the link. My current lineup is: 13 dBm source -> variable attenuator -> amp -> variable attenuator -> amp -> filter -> pad -> amp -> filter -> 20 dBm output I think I already paid attention to interleaving gain and loss, as well as not reducing the signal to very low levels so as to gain up the noise floor too much. At the moment I am also not much concerned with ALC behavior or stability as the AM noise measurements do not show much issue with this. Mostly I am concerned with additive noise from the somewhat complex opamp circuits driving (and linearizing) the series/shunt topology microwave FET voltage-variable attenuators. I would like to understand what changes in noise floor are just a result of attenuating the RF chain over it's amplitude control range, and what noise I am adding by my imperfect (noisy) attenuator driver circuitry. I have already made some improvements in this area by bandwidth-limiting the opamps (I need about 1 MHz 3 dB corner for ALC stability and AM modulation) and switching to some lower noise opamps. So clearly it was not perfect. Just trying to figure out how much is added AM noise by my circuit and how much is laws of physics by increasing the attenuation. Lifespeed On Wed, 1/29/14, John Miles wrote: Subject: Re: [time-nuts] help me understand AM noise To: "'Discussion of precise time and frequency measurement'" Date: Wednesday, January 29, 2014, 4:46 PM If you're just using a single attenuator at the input of your PA, it makes sense that the additive noise in dBc terms is worse at lower output levels. Disregarding saturation for the moment, the PA is adding just as much noise to a low-level input signal as it does to a high-level signal. ALC-controlled amps need to be designed with multiple alternating stages of attenuation and gain. Whatever drives the PIN diodes or other attenuators also needs to be quiet and well-filtered, of course -- which also means keeping an eye on feedback stability. My (least) favorite counterexample is the IF amp in the HP 11729C noise test set. It's not even an ALC amplifier, but a true limiting amplifier. It was implemented by cascading several high-gain MMICs with clamp diodes between each stage. Makes a nice comb generator. They were driving a mixer so they would have reasoned that odd harmonics were acceptable, but because there was no interstage attenuation for low-level noise, the amplifier's additive AM and PM levels were awful. You'll see similar performance if you measure the additive noise of a comparator or other limiting amp. Several limiting amps were used in the HP 8662A/8663A synthesizer as well, which is why they have good close-in noise but a relatively poor white noise floor. As long as you avoid these mistakes you'll be ahead of 90% of the crowd. As far as further reading goes, check out http://rubiola.org/index.html . -- john, KE5FX Miles Design LLC > -Original Message- > From: time-nuts-boun...@febo.com [mailto:time-nuts- > boun...@febo.com] On Behalf Of life speed > Sent: Wednesday, January 29, 2014 10:32 AM > To: time-nuts@febo.com > Subject: [time-nuts] help me understand AM noise > > Hi Guys. It has been a while since I posted, hope you can help with a slightly > time-related topic. Can't have frequency without amplitude . . . > > I recently designed an Automatic Level Control circuit consisting of dual- > slope detector logger, open and closed loop references with AM > modulation, and a linearizer (volts/dB) driver for series/shunt microwave > attenuators. This is part of a DC - 20 GHz microwave synthesizer. I measured > the AM noise at 3 GHz, both open and closed loop, and find the noise level > is higher at the output of the attenuator/amplifier chain at similar power > levels to the input (13 dBm). The input RF chain saturates at about 17 dBm, > while the output amp following the attenuators saturates at about 20 dBm. > > I understand that an amplifier in compression will suppress AM noise. What > I wonder is are my measurements of increased AM noise (red trace) at the > output of the attenuator/amp lineup to be expected based on the higher > available saturated power? Is it possible to attenuate the signal using the > power control (open loop in this example, ALC is not used) without > degrading AM noise performance? Does anybody have any suggested > reading on this subject? I am trying to understand how well my circuit > performs, in general. I do observe that control the power to a lower level > increases the AM noise. But it is a relative measurement to begin with, so > what is "good"? I have
[time-nuts] phase noise in digital divider, 2nd harmonic
Hi, Been a while since I visited, I recall there are many, well, time nuts here. I am trying to track down a source of phase noise in a frequency synthesizer design. One part of the frequency reference upconverts a DDS and then divides it down again using a digital divider - standard technique for DDS angle modulation spurious reduction. The DDS tunes over more than an octave, so obviously the single low pass filter isn't going to cut it. I am noticing up to 3 dB phase noise degradation at the output of the divider as the DDS frequency decreases and the 2nd harmonic is in-band to the LPF. I suspect this is disturbing the threshold crossing in the high speed digital logic divider, as described in "The Effect of Harmonic Distortion on Phase errors in Frequency Distribution and Synthesis" by F.L. Walls et al at NIST. What do you think? I should probably put in a switched filter to get my 3 dB back ;) - Lifespeed ___ 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] GPS discipline oscillator vs phase lock
I may already know the answer to this, but I figured I would ask the time nuts anyway. I have an application where I would like to phase-lock two oscillators together, probably 10MHz OCXOs as they have particularly good Allen Deviation compared to what I would ultimately like to use, a 100MHz crystal oscillator locally PLL'd to the 10MHz oscillators. These oscillators will be separated by a distance of a few yards up to a few miles. The requirement is not that their phases align perfectly, as in the conventional locally-connected phase-locked loop sense, but rather that any phase difference between the two oscillators resulting from arrival times of GPS signals are held constant. Perhaps this shows my lack of understanding of GPS time, I don't know if travel time is accounted for in commercial GPS 1Hz outputs, it may well be corrected. I suspect the result of a GPSDO is not the same as phase-locking two oscillators together. Perhaps it is frequency locking? Which, if the phase difference were held constant to within 0.1 degree, would be acceptable. Not sure this is the result either. Thanks for the info, - Lifespeed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS discipline oscillator vs phase lock
I was afraid of that, I guess it doesn't hurt to ask. Perhaps I could implement an ISM band radio link for the purpose of locking the two oscillators. Of course that wouldn't reach a couple miles either. - Lifespeed Hi A *lot* depends on your definition of “phase locked”. If indeed you are after 0.1 degree at 100 MHz, that gets into the “no can do” range. To put some numbers on it, 0.1 degree at 100 MHz is 2.7 ps. GPS time as received simply is not stable to that level … If you drop back to about 20 degrees, you start to get into the “maybe can do” range. 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] GPS discipline oscillator vs phase lock
That sounds like phase-locking the oscillators to a local radio transmitter. Not sure there is any difference post-processing vs. real time. - Lifespeed Hi A far more common approach is to let the two oscillators free run and to record something like a local broadcast station. You then post process all of the data to give you the phase accuracy. One of several gotcha’s is the stability of any radio link at the level you are looking for. 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.