Re: [time-nuts] HP 5370B jitter
Gentleman, I few years ago I bought an HP5370A about the seller was saying that he was selling it that cheap (200 US$) because he was not able to test whether this device would hold its specifications. As it turned it displayed SOMETIMES large amounts (a few hundred picoseconds) of RMS jitter and NOT on other times on the handbook's jitter self test. I have been trying a lot to improve the situation by a) correctly adjusting the input trigger circuits b) correctly adjusting the 200 MHz multiplier c) correctly adjusting the two interpolator boards Each of the adjustment procedures went exactly as descibed in the handbook. However, there was no improvement on the total situation. The counter insisted on sometimes behaving this way and sometimes the other way. Sometimes it seemed as if there were even a slow continouos movement between the two extremes perhaps due to temperature. I had already realized that the counter was scrap with no sale value when I remembered my friend Frank who happens to repair old HP stuff for one of the local surplus dealers. He had an defective 5370A around so that I was able to make some exchanges on board level. After changing some boards forward and back it became appearant that exchanging ONE of the interpolator boards would immediatly change the situation in that way that the counter would display an RMS jitter of 30-40 ps for any amount of time I would give him. So clearly the this interpolator board had an problem. Most of the electronic stuff on the interpolator board is plain electronics that might been gathered even today with the BIG exception of the oscillators that are part of the phase startable plls on the interpolator boards. These oscillators are of the delay line type and are contained in the unusual looking metal cans on the interpolator boards. So I had to test, whether the problem was located in this oscillator (in which case I was in difficulty because it it a HP propiarity product) or elsewhere (in which case I would be probably able to repair the board). By exchanging the oscillator between the boards the problem wandered, so clearly the problem WAS in the oscillator. Fortunately enough I managed to coax Frank a bad interpolator board with a good oscillator and since that my 5370 works ok. Best regads Ulrich Bangert -Ursprüngliche Nachricht- Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Im Auftrag von Didier Juges Gesendet: Samstag, 14. Juli 2007 05:38 An: 'Discussion of precise time and frequency measurement' Betreff: Re: [time-nuts] HP 5370B jitter ); SAEximRunCond expanded to false Errors-To: time-nuts-bounces+df6jb=ulrich-bangert.de+df6jb=ulrich-bangert [EMAIL PROTECTED] Magnus, My first 5370A had severe distortion on the 10 MHz output, but in my case. It looked like 30 MHz with some 10 MHz component (no 5 MHz), and was otherwise stable. It turned out some of the capacitors in the output filter of the amplifier board had cold solder joints (not unusual on these units), causing the tank circuit to resonate at 30 MHz instead of 10. Once reflowed, all went back to normal (except that did not fix the bad socket problem, but that's another issue :-) Didier KO4BB -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Magnus Danielson Sent: Friday, July 13, 2007 4:04 PM To: time-nuts@febo.com Subject: [time-nuts] HP 5370B jitter Hi fellow time-nuts! As my vacation started this evening, I spent some time in the lab (at work, since that's where I find my Wavecrest and it has better selection of active probes), trimming up the 200 MHz multiplier chain. I shifted it down from 17 ps RMS to 3.9-4.0 ps RMS. Most of the trimmings where on the 200 MHz part where as the 50 MHz part where fairly clean already from start. I may be able to tweak the jitter of the 200 MHz down further, but it would require more work than my rather quick-and-dirty approach. However, I was quite supprised to notice that there is alot of modulation on the 10 MHz output. The histogram shows two distinct gaussian bells and when checking the high-frequency modulation it showed a 5 MHz modulation. The spectrum analyzer clearly shows the 5 MHz output. I will make more investigation to the source of that 5 MHz, but it is annoying. If it is a property that my 5370B shares with others, then it will certainly be a limiting factor for self-referenced jitter measures as well as use of the output for low-jitter measures. I think the main source of sub-10 MHz clocks is the divide down on the CPU board A9. It generates control signals which is sent along the motherboard so there is alot of chance to jump over and infect the output buffer. Could you please have a look with a spectrum analyzer on your 10 MHz output. I have not checked how it behaves with external 10 MHz applied. Cheers, Magnus
Re: [time-nuts] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] From: [EMAIL PROTECTED] Subject: Re: [time-nuts] HP 5370B jitter Date: Fri, 13 Jul 2007 17:16:38 EDT Message-ID: [EMAIL PROTECTED] Said, Returning to this message... I sent spectrum plots of the 10MHz 5370B output to this list earlier, the output is very sad. It's extremely dirty and jittery (well, compared to the 4ps noise you got, and compared to what the 10811 can deliver). Further investigations have now shown that the 10811 output is clean, just a little 3rd harmonic but nothing to worry about. The INT test point is also clean. I suspect the output drive part, as other outputs to various parts are also free of the 5 MHz. The 10 MHz present detector circuit is currently my main suspect. It consists of a one-shot multi-vibrator triggered by the 10 MHz signal. Having a RC time-constant of 100 ns makes it a suspect indeed. Probing it (pin 11 on A8U1) clearly shows a waveform wich looks like a 25% 5 MHz with a short spike on it. This little culprit of a detector is infact a wideband comb-generator which contributes its 5 MHz as sidebands to the output 10 MHz. It's only purpose in life is to light a LED only visible to the servicing engineer (me in this case). Thus, making a small modification to disable it during normal operation would improve the quality of the 10 MHz output considerably if I am right. Since I don't do ECL design on a daily basis, I will have to ponder a bit in order to come up with a good method of acheiving this. After calibration to 4ps, what does your 5370B unit read for it's internal RMS noise now? Self-reference from 10 MHz output gives 29.9-34.0 ps. A 130 MHz +15 dBm sine results in readings in the 35-41 ps range. A 5 MHz +15 dBm sine from the Cesium does not acheive as deep levels, but reaches 39-43 ps but does not vary as much. A 10 MHz from anther 10811 source reaches 37-43 ps. All measures where taken in TI mode, +/- TI, StdDev for 1k samples. Optimum 1M/50 Ohm termination strategy selected for each signal. Flipping the input termination switches for both channels quickly creates different levels of input jitter. I have not trimmed anything but the multiplier stage. I have tested running frequency measurements and up to about 130 MHz seems safe. At 133 MHz it will start to false-trigger and at 134 MHz it is just starting to make obvious errors and at higher numbers it gives nonsense readings. I might add that I have only trimmed the 200 MHz multiplier chain so far. Cheers, Magnus ___ time-nuts mailing 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] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] Hej Magnus Magnus Danielson wrote: Further investigations have now shown that the 10811 output is clean, just a little 3rd harmonic but nothing to worry about. The INT test point is also clean. I suspect the output drive part, as other outputs to various parts are also free of the 5 MHz. The 10 MHz present detector circuit is currently my main suspect. It consists of a one-shot multi-vibrator triggered by the 10 MHz signal. Having a RC time-constant of 100 ns makes it a suspect indeed. Probing it (pin 11 on A8U1) clearly shows a waveform wich looks like a 25% 5 MHz with a short spike on it. This little culprit of a detector is infact a wideband comb-generator which contributes its 5 MHz as sidebands to the output 10 MHz. It's only purpose in life is to light a LED only visible to the servicing engineer (me in this case). Thus, making a small modification to disable it during normal operation would improve the quality of the 10 MHz output considerably if I am right. Since I don't do ECL design on a daily basis, I will have to ponder a bit in order to come up with a good method of acheiving this. Try shorting AU8U1 Pins 6 and 7 to ground. This turns of the npn output emitter followers driving the longtailed pair which drives the LED. This is permitted and nothing will be destroyed or degraded even if the shorts are permanent. Cheers, Magnus Bruce ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] From: Magnus Danielson [EMAIL PROTECTED] Subject: Re: [time-nuts] HP 5370B jitter Date: Sun, 15 Jul 2007 02:54:10 +0200 (CEST) Message-ID: [EMAIL PROTECTED] Hej fellow time-nuts, So, no more 5 MHz with overtones on my HP5370B thanks to this quick fix. Self-referenced jitter did not change significantly thought. It is actually slightly higher numbers, but that may be due to heatup or something. I noticed that during my previous run too. After carefull tuning of the input trigger level while still running +/- TI and 1k samples, I am now down to the 13.0-15.0 ps RMS range. The MEAN for those numbers is at -1.85 ns. This by itself should be an interesting little piece of knowledge and correlates well with what is to be expected. To acheive these numbers for a broader range of TI values one would have to work a bit on the instrument thought. It is nice to know that it can reach down there at least. Cheers, Magnus ___ time-nuts mailing 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] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] In a message dated 7/14/2007 16:45:19 Pacific Daylight Time, [EMAIL PROTECTED] writes: It's only purpose in life is to light a LED only visible to the servicing engineer (me in this case). Thus, making a small modification to disable it during normal operation would improve the quality of the 10 MHz output considerably if I am right. Since I don't do ECL design on a daily basis, I will have to ponder a bit in order to come up with a good method of acheiving this. Hi Magnus, yes, seems this circuit needs to go. Let me know how you disable it. Thanks, Said ** Get a sneak peak of the all-new AOL at http://discover.aol.com/memed/aolcom30tour ___ time-nuts mailing 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] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] In a message dated 7/13/2007 14:09:43 Pacific Daylight Time, [EMAIL PROTECTED] writes: probes), trimming up the 200 MHz multiplier chain. I shifted it down from 17 ps RMS to 3.9-4.0 ps RMS. Most of the trimmings where on the 200 MHz part where as the 50 MHz part where fairly clean already from start. Hi Magnus, I sent spectrum plots of the 10MHz 5370B output to this list earlier, the output is very sad. It's extremely dirty and jittery (well, compared to the 4ps noise you got, and compared to what the 10811 can deliver). After calibration to 4ps, what does your 5370B unit read for it's internal RMS noise now? bye, Said ** Get a sneak peak of the all-new AOL at http://discover.aol.com/memed/aolcom30tour ___ time-nuts mailing 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] HP 5370B jitter
I don't see much of anything around 5 MHz offsets (attached)... -- john, KE5FX However, I was quite supprised to notice that there is alot of modulation on the 10 MHz output. The histogram shows two distinct gaussian bells and when checking the high-frequency modulation it showed a 5 MHz modulation. The spectrum analyzer clearly shows the 5 MHz output. I will make more investigation to the source of that 5 MHz, but it is annoying. If it is a property that my 5370B shares with others, then it will certainly be a limiting factor for self-referenced jitter measures as well as use of the output for low-jitter measures. I think the main source of sub-10 MHz clocks is the divide down on the CPU board A9. It generates control signals which is sent along the motherboard so there is alot of chance to jump over and infect the output buffer. Could you please have a look with a spectrum analyzer on your 10 MHz output. I have not checked how it behaves with external 10 MHz applied. attachment: 5370.gif___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] From: [EMAIL PROTECTED] Subject: Re: [time-nuts] HP 5370B jitter Date: Fri, 13 Jul 2007 17:16:38 EDT Message-ID: [EMAIL PROTECTED] Hi Said, I sent spectrum plots of the 10MHz 5370B output to this list earlier, the output is very sad. It's extremely dirty and jittery (well, compared to the 4ps noise you got, and compared to what the 10811 can deliver). OK. So you have seen it too. I will ponder a little about if there is a cure for it. Fortunatly the Frequency Buffer Board is fairly easy to get to. I'll think I'll handgranade a hole on the lab-bench and pop the lid there. After calibration to 4ps, what does your 5370B unit read for it's internal RMS noise now? Still annoyingly high. Don't remember the actual number. It was jumping around like hell. I was not doing it properly thought, so I will redo it later tonight. I'll also see what kid of numbers I can get from good sources. There is plenty to pick from here at home. Cheers, Magnus ___ time-nuts mailing 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] HP 5370B jitter
); SAEximRunCond expanded to false Errors-To: [EMAIL PROTECTED] Magnus, My first 5370A had severe distortion on the 10 MHz output, but in my case. It looked like 30 MHz with some 10 MHz component (no 5 MHz), and was otherwise stable. It turned out some of the capacitors in the output filter of the amplifier board had cold solder joints (not unusual on these units), causing the tank circuit to resonate at 30 MHz instead of 10. Once reflowed, all went back to normal (except that did not fix the bad socket problem, but that's another issue :-) Didier KO4BB -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Magnus Danielson Sent: Friday, July 13, 2007 4:04 PM To: time-nuts@febo.com Subject: [time-nuts] HP 5370B jitter Hi fellow time-nuts! As my vacation started this evening, I spent some time in the lab (at work, since that's where I find my Wavecrest and it has better selection of active probes), trimming up the 200 MHz multiplier chain. I shifted it down from 17 ps RMS to 3.9-4.0 ps RMS. Most of the trimmings where on the 200 MHz part where as the 50 MHz part where fairly clean already from start. I may be able to tweak the jitter of the 200 MHz down further, but it would require more work than my rather quick-and-dirty approach. However, I was quite supprised to notice that there is alot of modulation on the 10 MHz output. The histogram shows two distinct gaussian bells and when checking the high-frequency modulation it showed a 5 MHz modulation. The spectrum analyzer clearly shows the 5 MHz output. I will make more investigation to the source of that 5 MHz, but it is annoying. If it is a property that my 5370B shares with others, then it will certainly be a limiting factor for self-referenced jitter measures as well as use of the output for low-jitter measures. I think the main source of sub-10 MHz clocks is the divide down on the CPU board A9. It generates control signals which is sent along the motherboard so there is alot of chance to jump over and infect the output buffer. Could you please have a look with a spectrum analyzer on your 10 MHz output. I have not checked how it behaves with external 10 MHz applied. Cheers, Magnus ___ time-nuts mailing 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.
