Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi Welcome to the world of trying to measure this stuff … > On Jan 9, 2015, at 6:53 AM, steph.rey wrote: > > Dear all, > > I'm trying to measure Alan Deviations using Timelab and some frequency > counters. > The device under test is a GPSDO using a TCXO as référence > > I've an HP58503B GPSDO which feeds my counters. I've tried an HP5342A, 0-18 > GHz, 1 Hz resolution and a Philipps PM6654C, 0.01Hz resolution. > > In Timelab, the plot with the HP5342A is around 10e-7 which correspond to 1Hz > and with the PM6654C, the plot is around 10e-10. > I would suspect that this is still the counter which limits the actual > response of my device under test. Yes, the counters and TCXO are limiting your measurements. > > My question are : > - how to measure Alan Deviations with levels below 10e-12/10e-13 ? How much money do you have to spend? ( There are expensive commercial ways to do this). No matter what, you will need a “better than” reference. That’s not going to be cheap. Most of us simply get a second GPSDO and compare them. The assumption is that they both are the same and you can allocate the error equally between them. With three you can more accurately allocate the error. A DMTD is the “cheap” way to get the actual measurement done. > What can be the application of measurement Alan deviation > 10e-10 ? I guess > most of the low frequency There are a number of systems applications that very much need good ADEV. Getting into why this or that nav or com system needs it would take a bit of time. > - The HP53503 GPS is given to be 10e-11 / 10e-12. I guess this will limit > anyway the measurement floor. I've a Rb source, but it's stability is within > the same range. What kind of reference would be more suitable for such > measurements ? If you want to do it directly, a hydrogen maser is a good way to go. That’s silly expensive. Just compare GPSDO’s, that’s a lot cheaper. > - With the PM6654C on 15h measurement, I can see some frequency jumps of 800 > Hz which are not relevants with the GPSDO undertest. I suspect error in data > transmission. This makes the overall measurement totally wrong (10e-5). The > counter is in talk only mode. I'd like to get rid of these points maybe 40-50 > points out of 1. Is there a way to do that from Timelab or the only > option is to export the file and process manually the data ? You can expand the data and zap the offending segments. It’s done on the phase plot. Have Fun. Bob > > Thanks & cheers > Stephane > ___ > time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hi Bob, Many thanks for your prompt and detailled answer. My question on applications wasn't on good ADEV where I perfetcly understand the need, but actually what could be the applications of measuring BAD ADEV (>10e-7). That was my point asking what king of application can we cover by measuring such high ADEV when you have counters with resolution not greater than 0.01Hz However you bring to me part of the answer when you talk about the reference and the way to get something cheap and better than 10e-12. I will investigate on DMTD. However, even if you have a beautiful Maser source, will you improve anything above the resolution of your counter. In other words, with my 0.01Hz counter, will I improve my measurement if I replace my GPSDO source with something much better ? I feel the resolution of the counter will anyway limit the ADEV floor, right ? If the last digit of the counter do not move how could we measure something smaller ? The counters I'm using are not running on their own reference (OCXO or TCXO) but with the HP58503b which is a GPS disciplined OCXO but with stability in the range of 10e-11 or 10e-12 at best. I'm working for a big lab where possibly I could have nice piece of equipment but this is always easier to find alternatives solutions at lower price. On the application I'm working on we're looking for phase stability in the range of fs at several GHz. One of the project I'm working will use a femtosecond laser modulated at 88 Mhz that some people want to use as RF reference for the 3 GHz source. I'm pretty sure this can't achieve the phase stability requirement and I'm trying to illustrate this. However even for my ham activites where I'm trying to design low noise LOs, I'd like to have a tool able to measure goog frequency and phase stability... Stephane On Fri, 9 Jan 2015 07:48:42 -0500, Bob Camp wrote: Hi Welcome to the world of trying to measure this stuff … On Jan 9, 2015, at 6:53 AM, steph.rey wrote: Dear all, I'm trying to measure Alan Deviations using Timelab and some frequency counters. The device under test is a GPSDO using a TCXO as référence I've an HP58503B GPSDO which feeds my counters. I've tried an HP5342A, 0-18 GHz, 1 Hz resolution and a Philipps PM6654C, 0.01Hz resolution. In Timelab, the plot with the HP5342A is around 10e-7 which correspond to 1Hz and with the PM6654C, the plot is around 10e-10. I would suspect that this is still the counter which limits the actual response of my device under test. Yes, the counters and TCXO are limiting your measurements. My question are : - how to measure Alan Deviations with levels below 10e-12/10e-13 ? How much money do you have to spend? ( There are expensive commercial ways to do this). No matter what, you will need a “better than” reference. That’s not going to be cheap. Most of us simply get a second GPSDO and compare them. The assumption is that they both are the same and you can allocate the error equally between them. With three you can more accurately allocate the error. A DMTD is the “cheap” way to get the actual measurement done. What can be the application of measurement Alan deviation > 10e-10 ? I guess most of the low frequency There are a number of systems applications that very much need good ADEV. Getting into why this or that nav or com system needs it would take a bit of time. - The HP53503 GPS is given to be 10e-11 / 10e-12. I guess this will limit anyway the measurement floor. I've a Rb source, but it's stability is within the same range. What kind of reference would be more suitable for such measurements ? If you want to do it directly, a hydrogen maser is a good way to go. That’s silly expensive. Just compare GPSDO’s, that’s a lot cheaper. - With the PM6654C on 15h measurement, I can see some frequency jumps of 800 Hz which are not relevants with the GPSDO undertest. I suspect error in data transmission. This makes the overall measurement totally wrong (10e-5). The counter is in talk only mode. I'd like to get rid of these points maybe 40-50 points out of 1. Is there a way to do that from Timelab or the only option is to export the file and process manually the data ? You can expand the data and zap the offending segments. It’s done on the phase plot. Have Fun. Bob Thanks & cheers Stephane ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
On 1/9/15 7:42 AM, steph.rey wrote: Hi Bob, Many thanks for your prompt and detailled answer. My question on applications wasn't on good ADEV where I perfetcly understand the need, but actually what could be the applications of measuring BAD ADEV (>10e-7). That was my point asking what king of application can we cover by measuring such high ADEV when you have counters with resolution not greater than 0.01Hz However you bring to me part of the answer when you talk about the reference and the way to get something cheap and better than 10e-12. I will investigate on DMTD. However, even if you have a beautiful Maser source, will you improve anything above the resolution of your counter. In other words, with my 0.01Hz counter, will I improve my measurement if I replace my GPSDO source with something much better ? I feel the resolution of the counter will anyway limit the ADEV floor, right ? If the last digit of the counter do not move how could we measure something smaller ? The various heterodyne techniques (DMTD is but one) let you use your counter for many more digits than it has. Essentially what you do is beat your unknown against the standard, and count the difference frequency. What you really do is put an offset in one (say 100Hz on a 10 MHz standard) so you're accurately measuring a 100 Hz instead of a 10 MHz signal. Your counter then gets down into the microHz. The other approach is to use one standard to drive the ADC clock to sample the unknown, and then post process in software. Once you've got a series of numbers, you can get infinite precision in software. there's a variety of schemes here too. The counters I'm using are not running on their own reference (OCXO or TCXO) but with the HP58503b which is a GPS disciplined OCXO but with stability in the range of 10e-11 or 10e-12 at best. I'm working for a big lab where possibly I could have nice piece of equipment but this is always easier to find alternatives solutions at lower price. On the application I'm working on we're looking for phase stability in the range of fs at several GHz. One of the project I'm working will use a femtosecond laser modulated at 88 Mhz that some people want to use as RF reference for the 3 GHz source. I'm pretty sure this can't achieve the phase stability requirement and I'm trying to illustrate this. However even for my ham activites where I'm trying to design low noise LOs, I'd like to have a tool able to measure goog frequency and phase stability... Stephane On Fri, 9 Jan 2015 07:48:42 -0500, Bob Camp wrote: Hi Welcome to the world of trying to measure this stuff … On Jan 9, 2015, at 6:53 AM, steph.rey wrote: Dear all, I'm trying to measure Alan Deviations using Timelab and some frequency counters. The device under test is a GPSDO using a TCXO as référence I've an HP58503B GPSDO which feeds my counters. I've tried an HP5342A, 0-18 GHz, 1 Hz resolution and a Philipps PM6654C, 0.01Hz resolution. In Timelab, the plot with the HP5342A is around 10e-7 which correspond to 1Hz and with the PM6654C, the plot is around 10e-10. I would suspect that this is still the counter which limits the actual response of my device under test. Yes, the counters and TCXO are limiting your measurements. My question are : - how to measure Alan Deviations with levels below 10e-12/10e-13 ? How much money do you have to spend? ( There are expensive commercial ways to do this). No matter what, you will need a “better than” reference. That’s not going to be cheap. Most of us simply get a second GPSDO and compare them. The assumption is that they both are the same and you can allocate the error equally between them. With three you can more accurately allocate the error. A DMTD is the “cheap” way to get the actual measurement done. What can be the application of measurement Alan deviation > 10e-10 ? I guess most of the low frequency There are a number of systems applications that very much need good ADEV. Getting into why this or that nav or com system needs it would take a bit of time. - The HP53503 GPS is given to be 10e-11 / 10e-12. I guess this will limit anyway the measurement floor. I've a Rb source, but it's stability is within the same range. What kind of reference would be more suitable for such measurements ? If you want to do it directly, a hydrogen maser is a good way to go. That’s silly expensive. Just compare GPSDO’s, that’s a lot cheaper. - With the PM6654C on 15h measurement, I can see some frequency jumps of 800 Hz which are not relevants with the GPSDO undertest. I suspect error in data transmission. This makes the overall measurement totally wrong (10e-5). The counter is in talk only mode. I'd like to get rid of these points maybe 40-50 points out of 1. Is there a way to do that from Timelab or the only option is to export the file and process manually the data ? You can expand the data and zap the offending segments. It’s done on the phase plot. Have Fun. Bob Tha
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi If your only instrument is a counter. — and — You never measure past 1x10^-10 with that counter — and — Measurements that bounce around with a standard deviation of the difference between readings of 1x10^-10 are ok. — then — No, you don’t need anything better than a 1x10^-10 ADEV. Most people would be bothered by a counter that has an typical jump of 1x10^-10 between every reading, so most would want a standard that’s a bit better than that. In addition, if you want to guarantee accuracy of a reading, you probably want something that’s 5X to 10X better than the level that stops the reading jitter. Simply put - ADEV is not standard deviation of frequency. Your frequency counter measures frequency. Going from one to the other means you want to have better ADEV than you might think. Bob > On Jan 9, 2015, at 10:42 AM, steph.rey wrote: > > Hi Bob, > > Many thanks for your prompt and detailled answer. > > My question on applications wasn't on good ADEV where I perfetcly understand > the need, but actually what could be the applications of measuring BAD ADEV > (>10e-7). That was my point asking what king of application can we cover by > measuring such high ADEV when you have counters with resolution not greater > than 0.01Hz > > However you bring to me part of the answer when you talk about the reference > and the way to get something cheap and better than 10e-12. I will investigate > on DMTD. However, even if you have a beautiful Maser source, will you improve > anything above the resolution of your counter. In other words, with my 0.01Hz > counter, will I improve my measurement if I replace my GPSDO source with > something much better ? I feel the resolution of the counter will anyway > limit the ADEV floor, right ? If the last digit of the counter do not move > how could we measure something smaller ? > The counters I'm using are not running on their own reference (OCXO or TCXO) > but with the HP58503b which is a GPS disciplined OCXO but with stability in > the range of 10e-11 or 10e-12 at best. > > I'm working for a big lab where possibly I could have nice piece of equipment > but this is always easier to find alternatives solutions at lower price. On > the application I'm working on we're looking for phase stability in the range > of fs at several GHz. One of the project I'm working will use a femtosecond > laser modulated at 88 Mhz that some people want to use as RF reference for > the 3 GHz source. I'm pretty sure this can't achieve the phase stability > requirement and I'm trying to illustrate this. > However even for my ham activites where I'm trying to design low noise LOs, > I'd like to have a tool able to measure goog frequency and phase stability... > > Stephane > > > > On Fri, 9 Jan 2015 07:48:42 -0500, Bob Camp wrote: >> Hi >> >> Welcome to the world of trying to measure this stuff … >> >>> On Jan 9, 2015, at 6:53 AM, steph.rey wrote: >>> >>> Dear all, >>> >>> I'm trying to measure Alan Deviations using Timelab and some frequency >>> counters. >>> The device under test is a GPSDO using a TCXO as référence >>> >>> I've an HP58503B GPSDO which feeds my counters. I've tried an HP5342A, 0-18 >>> GHz, 1 Hz resolution and a Philipps PM6654C, 0.01Hz resolution. >>> >>> In Timelab, the plot with the HP5342A is around 10e-7 which correspond to >>> 1Hz and with the PM6654C, the plot is around 10e-10. >>> I would suspect that this is still the counter which limits the actual >>> response of my device under test. >> >> Yes, the counters and TCXO are limiting your measurements. >> >>> >>> My question are : >>> - how to measure Alan Deviations with levels below 10e-12/10e-13 ? >> >> How much money do you have to spend? ( There are expensive commercial >> ways to do this). >> >> No matter what, you will need a “better than” reference. That’s not >> going to be cheap. Most of us simply get a second GPSDO and compare >> them. The assumption is that they both are the same and you can >> allocate the error equally between them. With three you can more >> accurately allocate the error. >> >> A DMTD is the “cheap” way to get the actual measurement done. >> >>> What can be the application of measurement Alan deviation > 10e-10 ? I >>> guess most of the low frequency >> >> There are a number of systems applications that very much need good >> ADEV. Getting into why this or that nav or com system needs it would >> take a bit of time. >> >>> - The HP53503 GPS is given to be 10e-11 / 10e-12. I guess this will limit >>> anyway the measurement floor. I've a Rb source, but it's stability is >>> within the same range. What kind of reference would be more suitable for >>> such measurements ? >> >> If you want to do it directly, a hydrogen maser is a good way to go. >> That’s silly expensive. Just compare GPSDO’s, that’s a lot cheaper. >> >>> - With the PM6654C on 15h measurement, I can see some frequency jumps of >>> 800 Hz which are not relevan
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Stephane, On 01/09/2015 12:53 PM, steph.rey wrote: Dear all, I'm trying to measure Alan Deviations using Timelab and some frequency counters. The device under test is a GPSDO using a TCXO as référence I've an HP58503B GPSDO which feeds my counters. I've tried an HP5342A, 0-18 GHz, 1 Hz resolution and a Philipps PM6654C, 0.01Hz resolution. In Timelab, the plot with the HP5342A is around 10e-7 which correspond to 1Hz and with the PM6654C, the plot is around 10e-10. I would suspect that this is still the counter which limits the actual response of my device under test. My question are : - how to measure Alan Deviations with levels below 10e-12/10e-13 ? What can be the application of measurement Alan deviation > 10e-10 ? I guess most of the low frequency - The HP53503 GPS is given to be 10e-11 / 10e-12. I guess this will limit anyway the measurement floor. I've a Rb source, but it's stability is within the same range. What kind of reference would be more suitable for such measurements ? - With the PM6654C on 15h measurement, I can see some frequency jumps of 800 Hz which are not relevants with the GPSDO undertest. I suspect error in data transmission. This makes the overall measurement totally wrong (10e-5). The counter is in talk only mode. I'd like to get rid of these points maybe 40-50 points out of 1. Is there a way to do that from Timelab or the only option is to export the file and process manually the data ? I've use the PM6654C with TimeLab. I wire the 10 MHz from the GPSDO and then the PPS to Channel A. Channel B has whatever signal I want to measure. By letting TimeLab know the frequency, it can adjust for any slipped cycles on the fly. This works well. The PM6654C has a single-shot resolution of 2 ns, which comes from the internal 500 MHz counting clock. This gives ~ 2E-9/tau (very coarse level) measurement limit. If you want to reach the 1E-12 resolution mark you need another 2000 of resolution gain, which is what you get if you mix your 10 MHz signal with a 10,005 MHz clock or lower. The Dual Mixer Time Difference (DMTD) is more likely to work well, as it provides some cancellation of the transfer clock. Slew-rates needs to be shaped, so you probably need a lower frequency to get some additional gain (and thus margin) and then amplifier stages on the beat signal. It's a tricky subject which requires a lot of attention to a bunch of details. I'd stay of the HP5342A as it will create dead-time in the measurements, which has a bias factor to it. The comments and suggestions you have received so far is also good comments. 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] question Alan deviation measured with Timelab and counters
Hi there, I'm still working on the ADEV measurement and here is what I've experimented today and few questions which are arising, that for sure have probably an easy answer... After having measured my 10 MHz GPSDOs with 1Hz and 0.01Hz resolution with HP8342A and Phillips PM6654C, I've seen the effect of resolution on the ADEV. With the 0.01 Hz resolution I could only achieve 1E-9 at 10 MHz which is very likely the floor limitation from the counter resolution. Actually I hope my GPSDO is better than that (~1E-12) The PM6654C can measure Time Intervals with a resolution of 0.01ps and an averaging time of 96s. I've then launch a new acquisition from Timelab, selecting Frequency difference instead of frequency only. I've not seen a time interval mode so I think this measurement gives nothing usable. The counter runs upon its own OCXO and I've connected two different GPSDO, one on each channel. I've tried both the internal gating as well as external gating with the 1PPS from the HP GPSDO with same result. The time interval is about 450ps with a variation of about 50ps. I've got an ADEV plot which is now in the range of 1E-12 / 1E-13. However here are my questions : - Can I use Frequency difference mode from Timelab to monitor time intervals ? If no is there a way to use the time interval measurement from the counter with Timelab to plot ADEV ? - In case the principle of plotting ADEV from Time Interval, what is the interpretation of the result ? The ADEV shows the relative stability between the two GPSDO... So, practically what does it bring ? And how to use this method if I want to characterize a device ? - I've googled for the DMTD and discovered the method. The principle seems clear and easy even if I know there might be several pitfails but here again, I don't know how to use the method or perform the result interpretation. In that method there is the LO and at leadt two DUTs inputs. I do understand that the method gives a relative stability between the two DUT... And what ? IF I've a device to characterize how can I get something useful by comparing with an other device that I do not know ? Where am I wrong here ? - stupid question on Timelab. If I let Timelab in Auto to select the period between two samples (correctly detected), the time scale of the graph is wrong. For instance, a 3h plot stops at 2000s (0.5h)... Here again, I miss something but what ? I'm sorry as it's probably weird questions already discussed many times but any comment or URL to point to would be a great help. Many thanks Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Bob Camp Envoyé : vendredi 9 janvier 2015 23:57 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi If your only instrument is a counter. — and — You never measure past 1x10^-10 with that counter — and — Measurements that bounce around with a standard deviation of the difference between readings of 1x10^-10 are ok. — then — No, you don’t need anything better than a 1x10^-10 ADEV. Most people would be bothered by a counter that has an typical jump of 1x10^-10 between every reading, so most would want a standard that’s a bit better than that. In addition, if you want to guarantee accuracy of a reading, you probably want something that’s 5X to 10X better than the level that stops the reading jitter. Simply put - ADEV is not standard deviation of frequency. Your frequency counter measures frequency. Going from one to the other means you want to have better ADEV than you might think. Bob > On Jan 9, 2015, at 10:42 AM, steph.rey wrote: > > Hi Bob, > > Many thanks for your prompt and detailled answer. > > My question on applications wasn't on good ADEV where I perfetcly > understand the need, but actually what could be the applications of > measuring BAD ADEV (>10e-7). That was my point asking what king of > application can we cover by measuring such high ADEV when you have > counters with resolution not greater than 0.01Hz > > However you bring to me part of the answer when you talk about the reference > and the way to get something cheap and better than 10e-12. I will investigate > on DMTD. However, even if you have a beautiful Maser source, will you improve > anything above the resolution of your counter. In other words, with my 0.01Hz > counter, will I improve my measurement if I replace my GPSDO source with > something much better ? I feel the resolution of the counter will anyway > limit the ADEV floor, right ? If the last digit of the counter do not move > how could we measure something smaller ? > The counters I'm using are not running on their own reference (OCXO or TCXO) > but with the HP5
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi Magnus, For some reason I've missed this message and the one from Jim until now ! This answers many of the questions I had. For my defense, I've 3000 messages since the last 3 months on the list !!! ah, yes, I'd like to get even better than 1E-12. 1E-14 would be perfect but my best standards for now are a HP GPSDO and an Effratrom FRK Rb which both are around 1E-12 'only'. I may have to invest in something better if prices are acceptable. I guess I won't be able to measure beyond the standard itself. The method you describes gives tau=2E-9 ? This is more or less what I could get with the frequency measurement (even a bit lower). So what is the benefit of the time interval measurement here against the frequency measurement ? However if I hear what you says, the GPSDO provides the 10 MHz standard reference for the counter, the GPSDO PPS on channel A and channel B receives for instance a 10 MHz signal I want to measure. So what will be the result of Time A-B then ? I do not understand why you put the PPS on channel A instead of something of the same frequency than the DUT ? How the time A-B will behave with these two different frequencies... " By letting TimeLab know the frequency, it can adjust for any slipped cycles on the fly." I guess this is what I've not understood. Now if I mix down the 10 MHz DUT with a 10.005 reference to increase the resolution, I'll get 5 kHz on channel B and still PPS on channel A ? Again I do not understand what will happen with these two signals on the time A-B. If I push your method a bit more, I could even get a beat frequency of 1 Hz and with 10-digits I would have increased my resolution by 10E6. Then I will be limited by the standard stability but on the principle would it work as well ? On that document http://www2.nict.go.jp/aeri/sts/2009TrainingProgram/Time%20Keeping/091017_DMTD.pdf it says (page 6) the accuracy of measurement is improved by a factor v/vb (the DUT and offset LO 1/2.PI.f). So it sounds to me that there is a compromise between resolution increase and accuracy. If I chose a beat frequency of 1 Hz the accuracy will not be improved but the resolution will be, right ? What is the transfer clock you're talking about ? and by the way should the offset LO be as stable as the standard reference meaning greater than the DUT ? Well, it's far too late here to let my brain working anymore. I will perform further experiments tomorrow at the office. Thanks & cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Magnus Danielson Envoyé : samedi 10 janvier 2015 02:05 À : time-nuts@febo.com Cc : mag...@rubidium.se Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Stephane, On 01/09/2015 12:53 PM, steph.rey wrote: > Dear all, > > I'm trying to measure Alan Deviations using Timelab and some frequency > counters. > The device under test is a GPSDO using a TCXO as référence > > I've an HP58503B GPSDO which feeds my counters. I've tried an HP5342A, > 0-18 GHz, 1 Hz resolution and a Philipps PM6654C, 0.01Hz resolution. > > In Timelab, the plot with the HP5342A is around 10e-7 which correspond > to 1Hz and with the PM6654C, the plot is around 10e-10. > I would suspect that this is still the counter which limits the actual > response of my device under test. > > My question are : > - how to measure Alan Deviations with levels below 10e-12/10e-13 ? > What can be the application of measurement Alan deviation > 10e-10 ? I > guess most of the low frequency > - The HP53503 GPS is given to be 10e-11 / 10e-12. I guess this will > limit anyway the measurement floor. I've a Rb source, but it's > stability is within the same range. What kind of reference would be > more suitable for such measurements ? > - With the PM6654C on 15h measurement, I can see some frequency jumps > of > 800 Hz which are not relevants with the GPSDO undertest. I suspect > error in data transmission. This makes the overall measurement totally > wrong (10e-5). The counter is in talk only mode. I'd like to get rid > of these points maybe 40-50 points out of 1. Is there a way to do > that from Timelab or the only option is to export the file and process > manually the data ? I've use the PM6654C with TimeLab. I wire the 10 MHz from the GPSDO and then the PPS to Channel A. Channel B has whatever signal I want to measure. By letting TimeLab know the frequency, it can adjust for any slipped cycles on the fly. This works well. The PM6654C has a single-shot resolution of 2 ns, which comes from the internal 500 MHz counting clock. This gives ~ 2E-9/tau (very coarse level) measurement limit. If you want to reach the 1E-12 resolution mark you need another 2000 of resolutio
Re: [time-nuts] question Alan deviation measured with Timelab and counters
> - Can I use Frequency difference mode from Timelab to monitor time intervals ? > If no is there a way to use the time interval measurement from the counter > with > Timelab to plot ADEV ? If you feed in frequency samples, it will convert them to phase-difference samples internally, so the program itself doesn't really care. The use of frequency data has a few drawbacks such as less accurate ADEV plots due to the counter's dead time between readings, but it's the easiest way to get started and is perfectly usable for many purposes. In general you should avoid letting the counter do any averaging. Except in very specific circumstances, any apparent improvement in ADEV measurement floor will be illusory. There are exceptions, but this isn't something you want to mess with until you're very comfortable with the rest of the measurement process. Your counter's true ADEV measurement floor at t=1s should be assumed to be close to its single-shot resolution specification (e.g., 100 ps = about 1E-10). > - In case the principle of plotting ADEV from Time Interval, what is the > interpretation of the result ? The ADEV shows the relative stability between > the > two GPSDO... So, practically what does it bring ? And how to use this method > if I > want to characterize a device ? An ADEV graph shows frequency stability statistics at different intervals, ranging from the rate at which the readings are returned from the counter (tau zero, at the left end of the plot) to a maximum interval that's related to how long you let the measurement run. It's much too deep a subject to go into in an email; see http://www.ke5fx.com/stability.htm for more pointers. Again, TimeLab always plots ADEV from time interval/phase data, even if you give it frequency readings. ADEV is fundamentally a frequency stability metric, but it can be computed identically from either TI or frequency samples (assuming zero dead time). > - stupid question on Timelab. If I let Timelab in Auto to select the period > between two samples (correctly detected), the time scale of the graph is > wrong. > For instance, a 3h plot stops at 2000s (0.5h)... Here again, I miss something > but > what ? The TimeLab manual, for one thing. :) Hit the books (specifically http://www.miles.io/TimePod_5330A_user_manual.pdf , page 31). -- john, KE5FX Miles Design LLC ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hi John, I hadn't noticed before you were here as well ;-) Thanks for answering. So I do understand I can use Timelab in frequency difference even if my counter sends data in TI in nanoseconds. Great. Ah and thanks for the manual link. I didn't remember this was in the manual of the Timepod Will investigate further today -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de John Miles Envoyé : mercredi 14 janvier 2015 07:26 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters > - Can I use Frequency difference mode from Timelab to monitor time intervals ? > If no is there a way to use the time interval measurement from the > counter with Timelab to plot ADEV ? If you feed in frequency samples, it will convert them to phase-difference samples internally, so the program itself doesn't really care. The use of frequency data has a few drawbacks such as less accurate ADEV plots due to the counter's dead time between readings, but it's the easiest way to get started and is perfectly usable for many purposes. In general you should avoid letting the counter do any averaging. Except in very specific circumstances, any apparent improvement in ADEV measurement floor will be illusory. There are exceptions, but this isn't something you want to mess with until you're very comfortable with the rest of the measurement process. Your counter's true ADEV measurement floor at t=1s should be assumed to be close to its single-shot resolution specification (e.g., 100 ps = about 1E-10). > - In case the principle of plotting ADEV from Time Interval, what is > the interpretation of the result ? The ADEV shows the relative > stability between the two GPSDO... So, practically what does it bring > ? And how to use this method if I want to characterize a device ? An ADEV graph shows frequency stability statistics at different intervals, ranging from the rate at which the readings are returned from the counter (tau zero, at the left end of the plot) to a maximum interval that's related to how long you let the measurement run. It's much too deep a subject to go into in an email; see http://www.ke5fx.com/stability.htm for more pointers. Again, TimeLab always plots ADEV from time interval/phase data, even if you give it frequency readings. ADEV is fundamentally a frequency stability metric, but it can be computed identically from either TI or frequency samples (assuming zero dead time). > - stupid question on Timelab. If I let Timelab in Auto to select the > period between two samples (correctly detected), the time scale of the graph > is wrong. > For instance, a 3h plot stops at 2000s (0.5h)... Here again, I miss > something but what ? The TimeLab manual, for one thing. :) Hit the books (specifically http://www.miles.io/TimePod_5330A_user_manual.pdf , page 31). -- john, KE5FX Miles Design LLC ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.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] question Alan deviation measured with Timelab and counters
Bonjour Stéphane, On 01/14/2015 02:16 AM, Stéphane Rey wrote: Hi Magnus, For some reason I've missed this message and the one from Jim until now ! This answers many of the questions I had. For my defense, I've 3000 messages since the last 3 months on the list !!! ah, yes, I'd like to get even better than 1E-12. 1E-14 would be perfect but my best standards for now are a HP GPSDO and an Effratrom FRK Rb which both are around 1E-12 'only'. I may have to invest in something better if prices are acceptable. I guess I won't be able to measure beyond the standard itself. The method you describes gives tau=2E-9 ? This is more or less what I could get with the frequency measurement (even a bit lower). So what is the benefit of the time interval measurement here against the frequency measurement ? I've been sloppy with the scaling factor, so there is a fixed scaling factor for the noise that the single-shot resolution produces, and that would be a measurement limit that if everything else is ideal would dominate. This quantization noise is sqrt(1/12) or about 0.289 if I remember correctly, so that is the scale-factor. It will also have a 1/tau slope. So that is how you can expect this noise to behave, it will look like white phase noise, but isn't, it is highly systematic noise, and if you play nicely with it, you can measure below it. However, doing so is non-trivial. I have one counter that does that. The good old HP5328A with the Option 040-series of boards will introduce noise to the counting 100 MHz oscillator such that averaging gets you down towards 10 ps rather than 10 ns resolution in TI mode. However, it does not help you to get nice frequency or stability measures. I've not taken the time to detail-analyse the ADEV scaling factor thought, I should do that, but it follows the general formula of ADEV(tau) = k*t_res/tau where t_res is the single-shot resolution and k is a constant. There is more to this, as counters can show up non-linearities of several sorts, and that the trigger conditions of the input has been optimized, which can be slew-rate limited for many counters and conditions. So, anyway, there is a bit of hand-waving in there, but I thought it was better to get you to "get" the basic trend there first, and then we can discuss the detailed numbers, as theory is one thing and achieved number can be quite a different one. As for frequency and time-interval measurements, if properly done, they can be used interchangeably without much impact. Realize that frequency and time-interval measurements will both be based on time-interval measurements as the core observation inside the counter, so the single-shot resolution limit applies to them both. However, subtle details lies in how the counter works and there is ways that the frequency precision can be lost. A good counter is the SR620, but the way it does the frequency measure, you need to calibrate the internal delay to make it "on the mark" measure. Using it in time-interval mode and you can eliminate that offset, because the start and stop measure of your signal under test is done with the same channel, with essentially the same delay both trigger-times. Another subtle detail is that when you make frequency measurements, you arm your counter, the start channel triggers, you wait the time you have programmed as the measurement time before you arm the stop channel, and then it triggers, after which you then read out your coarse counter of cycles, the interpolator states for the start and stop channels and well, the count of the time-base (which should be known), you calculate the frequency and output and well, once you cleared the "bench" from that measure you then arm the counter core of the next measurement. The time from the stop event to the following start event is called the dead-time. This dead-time is a period when the signal is not being observed. The actual time between the measures (time between the start events) and the length of the measures (time between the start and stop events) will not be the same, this will create a measurement bias in the ADEV. If you can establish the length of the dead-time you can compensate the measures. Very very few people do this these days, part of it is ignorance, part of it is why bother when you can use any of a number of techniques that avoid the dead-time altogether. Being able to measure frequency does not easily convert into making quality ADEV measures. Also, another danger of using frequency measures is that many modern counters use one of several techniques to improve the frequency measurement resolution by using things like linear regression. This behaves as a narrow-band filter, and the ADEV measures for white noise depend on the bandwidth of the system, and well, very very few measurements is annotated with their bandwidth, so traceable ADEV measurements will not be done there, and this pre-filtering effect band
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Bonjour Magnus, Many thanks for your very long and detailed answer. I've read quickly bu will go deeper tonight. Here are the results of today experiments... which are not giving anything valuable... I still don't understand the results I get :-/ With the PM6654C, I've put the HP GPSDO on the standard input, the 1 PPS on channel A and the 10 MHz from the DUT (GPSDO as well) on channel B. This gives something in the range of 2E-9 which looks like the counter resolution, right ? The gating takes 4s and the Time A-B displays a value like 64 E-6 Now if I downmix the channel B to 5 kHz (LO is a DDS Standford & Reseach generator), I have a sinus with lower amplitude and no squarer in my hand at the moment to shape the signal. Anyway, I do the same operation and I get on the display two more digits like xx.xx E-6 but the ADEC is in the range of E-7 I do not understand at all this fact. Even if the slew rate is not great, I was expecting an improvement. Note that the values displayed are always changing quite a lot between two samples. For instance with the 5 KHz channel B signal, I can read first sample at 27.11E-6, in the next one is 31.22E-6... which sounds huge, right ? I've then found an HP5370A and tried the same operation. Unfortunately the 5 kHz output is too low for the HP5370A sensitivity. I need an amplifier or sqauerer here but had no time to build on today. Si I could not get anything valuable with the HP5370A at the moment... Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Magnus Danielson Envoyé : mercredi 14 janvier 2015 08:04 À : time-nuts@febo.com Cc : mag...@rubidium.se Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Bonjour Stéphane, On 01/14/2015 02:16 AM, Stéphane Rey wrote: > Hi Magnus, > > For some reason I've missed this message and the one from Jim until now ! > This answers many of the questions I had. For my defense, I've 3000 messages > since the last 3 months on the list !!! > > ah, yes, I'd like to get even better than 1E-12. 1E-14 would be perfect but > my best standards for now are a HP GPSDO and an Effratrom FRK Rb which both > are around 1E-12 'only'. I may have to invest in something better if prices > are acceptable. I guess I won't be able to measure beyond the standard itself. > > The method you describes gives tau=2E-9 ? This is more or less what I could > get with the frequency measurement (even a bit lower). So what is the benefit > of the time interval measurement here against the frequency measurement ? I've been sloppy with the scaling factor, so there is a fixed scaling factor for the noise that the single-shot resolution produces, and that would be a measurement limit that if everything else is ideal would dominate. This quantization noise is sqrt(1/12) or about 0.289 if I remember correctly, so that is the scale-factor. It will also have a 1/tau slope. So that is how you can expect this noise to behave, it will look like white phase noise, but isn't, it is highly systematic noise, and if you play nicely with it, you can measure below it. However, doing so is non-trivial. I have one counter that does that. The good old HP5328A with the Option 040-series of boards will introduce noise to the counting 100 MHz oscillator such that averaging gets you down towards 10 ps rather than 10 ns resolution in TI mode. However, it does not help you to get nice frequency or stability measures. I've not taken the time to detail-analyse the ADEV scaling factor thought, I should do that, but it follows the general formula of ADEV(tau) = k*t_res/tau where t_res is the single-shot resolution and k is a constant. There is more to this, as counters can show up non-linearities of several sorts, and that the trigger conditions of the input has been optimized, which can be slew-rate limited for many counters and conditions. So, anyway, there is a bit of hand-waving in there, but I thought it was better to get you to "get" the basic trend there first, and then we can discuss the detailed numbers, as theory is one thing and achieved number can be quite a different one. As for frequency and time-interval measurements, if properly done, they can be used interchangeably without much impact. Realize that frequency and time-interval measurements will both be based on time-interval measurements as the core observation inside the counter, so the single-shot resolution limit applies to them both. However, subtle details lies in how the counter works and there is ways that the frequency precision can be lost. A good counter is the SR620, but the way it does the frequency measure, you need to calibrate the internal delay to make it "on the mark" measure. Using it in time-interval mode and you can eliminate that of
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Bonsoir Stéphane, (For Stéphane's unedited reply, look in the Current state thread) On 01/17/2015 12:36 AM, Stéphane Rey wrote: > Hi, > > I've took the time to read carefully your long and detailed message Magnus and this was very interesting. I've learned many things that have enabled me to investigate further. Ah yes, you're right saying that the more you fall into these things, the more you discover that you have to learn. Recently I've worked a lot on PLLs and I've actually learned a lot on special care to ensure low noise Very interesting. By the way I'm still working on this topic to improve again the noise (currently on a 3 GHz LO) Good that you found it useful. > Here are some experiment results : http://www.ptp-images.com/affiche-directement-l-image-kccsz71c9a.html > > 1. Setup #1 (blue plot) > > HP5370A > > standard input from HP GPSDO > > EXT input not connected, internal Arming 0.4s rate > > START input from 10 MHz distribution unit RacalDana 9478 Rubidium > > 2. Setup #2 (pink plot) > > HP5370A > > standard input from HP GPSDO > > EXT input not connected, internal Arming 0.4s rate > > START input from DUT (10 MHz homemade GPSDO) > > I'm not sure this is the proper way to connect everything... but this is the setup providing the lowest ADEV... which is between 1E-10 and 1E-13. But is the truth ? > > I feel strange the two plots having the same decreasing path along a linear slope (I mean linear on the log-log plot) ... I'm not sure of what I'm measuring ? Could this be the system measurement floor ? By the way how to measure the ADEV floor of a system other than having a source greater than the measurement system ? The slope is to be expected from the ADEV matching white PM noise, because it behaves similar enough to it on the ADEV plot. Please see the Allan Deviation wikipedia page, where I amongst other things added the power-law noise section with a handy table. https://en.wikipedia.org/wiki/Allan_variance > What could be these oscillations on the homemade (not by me) GPSDO ? Could be. Do use the phase plot (press p) to see if you have visible wobblings or something. > I've tried to downmix the DUT 10 MHz to few kHz using a SR DDS generator and a double balanced mixer from minicircuit via a low pass filter tuned at 100 kHz, but the level wasn't high enough for the counter (which I found strange as it was already nearly 200mV). I hadn't anything in hands to make a squarer quickly so I've just added a Minicircuits RF amplifier. The level was good but the ADEV has jumped to 1E-6. The signal was noisy already on the oscilloscope which I know is for sure the cause. Did you set the original frequency to 10 MHz in TimeLab? As you mix-down, TimeLab needs to be told what the original frequency was, it will calculate the gain and adjust the scale accordingly. Check your Acquire window. > I need to make a squarer. I was hesitating between several methods : using a CMOS gate, but this will increase the flicker noise from what I've read, using an amplifier and clamping diodes or a fast comparator which might create some noise around the trigger point... Any recommendation there ? Check out the Wenzel two-transistor design: http://www.wenzel.com/documents/waveform.html Buffer the output with an inverter or so. > I'll try to make this squarer next week to continue my investigations The joy of lab sessions. 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] question Alan deviation measured with Timelab and counters
Hi I’m a little concerned about the measurement setup here. Based on the quoted text, there have been a few messages in the thread that have not showed up here. = The “start” input on the counter is defined below. The “stop” input is not defined. Is the counter running in time interval mode or in frequency mode? IF it’s in time mode - what is the stop hooked to? IF it’s in frequent mode - what is the gate time set to ? Is the “standard input” the missing front panel input or is it the external reference input on the back panel? = I’m looking at the data on the link: http://www.ptp-images.com/affiche-directement-l-image-kccsz71c9a.html In both cases the slope is roughly 1/tau. Both plots end up in the 3 to 4 x 10^-13 range at 100 to 300 seconds. That’s suspiciously good performance for a rubidium or a GPSDO. Which is what makes me wonder about the setup. The Blue plot (1 pps?) ran for 18 minutes and has 1,114 points in it. The Pink plot ran for about 9 minutes and has a bit over 500 points in it. Both seem reasonable for a 1 pps to 10 MHz sort of setup. That may explain part of my confusion above. Again - I apologize if this all got explained in a post that went missing here. Bob > On Jan 17, 2015, at 7:34 PM, Magnus Danielson > wrote: > > Bonsoir Stéphane, > > (For Stéphane's unedited reply, look in the Current state thread) > > On 01/17/2015 12:36 AM, Stéphane Rey wrote: > > Hi, > > > > I've took the time to read carefully your long and detailed message Magnus > > and this was very interesting. I've learned many things that have enabled > > me to investigate further. Ah yes, you're right saying that the more you > > fall into these things, the more you discover that you have to learn. > > Recently I've worked a lot on PLLs and I've actually learned a lot on > > special care to ensure low noise Very interesting. By the way I'm still > > working on this topic to improve again the noise (currently on a 3 GHz LO) > > Good that you found it useful. > > > Here are some experiment results : > > http://www.ptp-images.com/affiche-directement-l-image-kccsz71c9a.html > > > > 1. Setup #1 (blue plot) > > > > HP5370A > > > > standard input from HP GPSDO > > > > EXT input not connected, internal Arming 0.4s rate > > > > START input from 10 MHz distribution unit RacalDana 9478 Rubidium > > > > 2. Setup #2 (pink plot) > > > > HP5370A > > > > standard input from HP GPSDO > > > > EXT input not connected, internal Arming 0.4s rate > > > > START input from DUT (10 MHz homemade GPSDO) > > > > I'm not sure this is the proper way to connect everything... but this is > > the setup providing the lowest ADEV... which is between 1E-10 and 1E-13. > > But is the truth ? > > > > I feel strange the two plots having the same decreasing path along a linear > > slope (I mean linear on the log-log plot) ... I'm not sure of what I'm > > measuring ? Could this be the system measurement floor ? By the way how to > > measure the ADEV floor of a system other than having a source greater than > > the measurement system ? > > The slope is to be expected from the ADEV matching white PM noise, because it > behaves similar enough to it on the ADEV plot. > Please see the Allan Deviation wikipedia page, where I amongst other things > added the power-law noise section with a handy table. > https://en.wikipedia.org/wiki/Allan_variance > > > What could be these oscillations on the homemade (not by me) GPSDO ? > > Could be. Do use the phase plot (press p) to see if you have visible > wobblings or something. > > > I've tried to downmix the DUT 10 MHz to few kHz using a SR DDS generator > > and a double balanced mixer from minicircuit via a low pass filter tuned at > > 100 kHz, but the level wasn't high enough for the counter (which I found > > strange as it was already nearly 200mV). I hadn't anything in hands to make > > a squarer quickly so I've just added a Minicircuits RF amplifier. The level > > was good but the ADEV has jumped to 1E-6. The signal was noisy already on > > the oscilloscope which I know is for sure the cause. > > Did you set the original frequency to 10 MHz in TimeLab? > As you mix-down, TimeLab needs to be told what the original frequency was, it > will calculate the gain and adjust the scale accordingly. > Check your Acquire window. > > > I need to make a squarer. I was hesitating between several methods : using > > a CMOS gate, but this will increase the flicker noise from what I've read, > > using an amplifier and clamping diodes or a fast comparator which might > > create some noise around the trigger point... Any recommendation there ? > > Check out the Wenzel two-transistor design: > http://www.wenzel.com/documents/waveform.html > > Buffer the output with an inverter or so. > > > I'll try to make this squarer next week to continue my investigations > > The joy of lab sessions. > > Cheers, > Magnus > _
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Stephane wrote: I need to make a squarer. I was hesitating between several methods : using a CMOS gate, but this will increase the flicker noise from what I've read, using an amplifier and clamping diodes or a fast comparator which might create some noise around the trigger point... Any recommendation there ? The basic Wenzel design can be optimized with attention to a few details (see below). Note that bias must be supplied to the LM329 (or whatever reference you use for the bases), which is not shown. The 20v supply should be quiet. The 1k emitter resistors can be replaced with quiet current sources, if you want to go to the trouble -- but with a quiet 20v supply, that should not be necessary (and designing quiet current sources is its own challenge). I posted a variant of this circuit for 3v logic on 12/25/14, which you can find in the archive. The power supply decoupling shown in that version (100uF tantalum + 1uF X7R ceramic + 10nF C0G or X7R ceramic) is superior to what is shown below). Best regards, Charles ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hi Bob, On 01/18/2015 04:25 AM, Bob Camp wrote: Hi I’m a little concerned about the measurement setup here. Based on the quoted text, there have been a few messages in the thread that have not showed up here. The messages got accidentally posted in the wrong thread. = The “start” input on the counter is defined below. The “stop” input is not defined. Is the counter running in time interval mode or in frequency mode? IF it’s in time mode - what is the stop hooked to? IF it’s in frequent mode - what is the gate time set to ? Is the “standard input” the missing front panel input or is it the external reference input on the back panel? I was also considering the setup strange in this regard. Common switched in? = I’m looking at the data on the link: http://www.ptp-images.com/affiche-directement-l-image-kccsz71c9a.html In both cases the slope is roughly 1/tau. Both plots end up in the 3 to 4 x 10^-13 range at 100 to 300 seconds. That’s suspiciously good performance for a rubidium or a GPSDO. Which is what makes me wonder about the setup. The Blue plot (1 pps?) ran for 18 minutes and has 1,114 points in it. The Pink plot ran for about 9 minutes and has a bit over 500 points in it. Both seem reasonable for a 1 pps to 10 MHz sort of setup. That may explain part of my confusion above. Again - I apologize if this all got explained in a post that went missing here. Not really. The plots looks to me like measurement setup baseline plots, with some sine noise in them. 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] question Alan deviation measured with Timelab and counters
Hi Ok, I didn’t think I’d seen the plots before. I agree that the plots look like “counter limited” data. That’s a fine explanation at the shorter Tau’s. I also agree that some sort of periodic “stuff” is getting into one of the signals and creating the ripple. What I’m wondering about (and what makes me question the setup) is the fact that the data is still “counter limited” at the mid to low parts in 10^-13 level at just a bit over 100 seconds. A telecom Rb is doing pretty well to be at 1x10^-12 at 100 seconds. Most GPSDO’s are doing well to be mid parts in 10^-12 at that tau. Simply put, the data continues to be counter limited to a pretty low point. Bob > On Jan 18, 2015, at 7:13 AM, Magnus Danielson > wrote: > > Hi Bob, > > On 01/18/2015 04:25 AM, Bob Camp wrote: >> Hi >> >> I’m a little concerned about the measurement setup here. Based on the quoted >> text, there have been a few messages in the thread that have not showed up >> here. > > The messages got accidentally posted in the wrong thread. > >> = >> >> The “start” input on the counter is defined below. The “stop” input is not >> defined. Is the counter running in time interval mode or in frequency mode? >> >> IF it’s in time mode - what is the stop hooked to? >> >> IF it’s in frequent mode - what is the gate time set to ? >> >> Is the “standard input” the missing front panel input or is it the external >> reference input on the back panel? > > I was also considering the setup strange in this regard. Common switched in? > >> = >> >> I’m looking at the data on the link: >> >> http://www.ptp-images.com/affiche-directement-l-image-kccsz71c9a.html >> >> In both cases the slope is roughly 1/tau. Both plots end up in the 3 to 4 x >> 10^-13 range at 100 to 300 seconds. That’s suspiciously good performance >> for a rubidium or a GPSDO. Which is what makes me wonder about the setup. >> >> The Blue plot (1 pps?) ran for 18 minutes and has 1,114 points in it. The >> Pink plot ran for about 9 minutes and has a bit over 500 points in it. Both >> seem reasonable for a 1 pps to 10 MHz sort of setup. That may explain part >> of my confusion above. >> >> Again - I apologize if this all got explained in a post that went missing >> here. > > Not really. The plots looks to me like measurement setup baseline plots, with > some sine noise in them. > > 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.
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hello, First, please do apologize for the confusion answering in the bad email. That's things I'm absolutely able to do when replying at 3 am ! Again, sorry for that and thanks Magnus for having corrected this. Back to my setup : There is indeed nothing on the STOP input of the HP5370a. The standard 10 MHz comes from the GPSDO HP-58503B and feeds the HP5370a Standard input. Its ADEV is given on page 240 of that document : http://www.leapsecond.com/museum/hp58503a/097-58503-13-iss-1.pdf We see that the shape is starting at about 2E-12 at 1s, increase to 2E-11 at 100s before decreasing again down to E-13 for above 10E3s... The setup #1 was using the Racal DANA Rb connected on the START input which is specified at E-9 / E-10, given on page 16 of the manual : http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Racal/9470-9479.pdf The EXT input receives the 1PPS from the HP58503b. It apparently drives both the START and the STOP of the acquisition (the two lights are blinking and the time between two measurement is no longer adjustable from the front panel RATE potentiometer and the period between two samples is 1.0s (detected by Timelab). But yes, the ADEV plot sounds really strange as it goes incredibly low after few seconds which is not consistent with the stability of the sources I'm using which is why I felt something was wrong On Setup #2 I've only replaced the Racal Dana Rb with the GPSDO to test. I've not made this design and not checked yet anything on it. Could these oscillations be from power supply noise ? To be checked. But how can it follow the ADEV plot of the Racal Dana Rb ? mmm Coincidence is not something I like too much and I believe something is clearly wrong in my measurement But what ??? On the Timelab setup screen before launching the acquisition I've left all the parameters as it without touching them. I've just seize 10E6 in the frequency field. Ah, by thay Magnus, for the downmixed test I've forgotten to change this value, I will check on monday when back at the office. Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Bob Camp Envoyé : dimanche 18 janvier 2015 14:44 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi Ok, I didn’t think I’d seen the plots before. I agree that the plots look like “counter limited” data. That’s a fine explanation at the shorter Tau’s. I also agree that some sort of periodic “stuff” is getting into one of the signals and creating the ripple. What I’m wondering about (and what makes me question the setup) is the fact that the data is still “counter limited” at the mid to low parts in 10^-13 level at just a bit over 100 seconds. A telecom Rb is doing pretty well to be at 1x10^-12 at 100 seconds. Most GPSDO’s are doing well to be mid parts in 10^-12 at that tau. Simply put, the data continues to be counter limited to a pretty low point. Bob > On Jan 18, 2015, at 7:13 AM, Magnus Danielson > wrote: > > Hi Bob, > > On 01/18/2015 04:25 AM, Bob Camp wrote: >> Hi >> >> I’m a little concerned about the measurement setup here. Based on the quoted >> text, there have been a few messages in the thread that have not showed up >> here. > > The messages got accidentally posted in the wrong thread. > >> = >> >> The “start” input on the counter is defined below. The “stop” input is not >> defined. Is the counter running in time interval mode or in frequency mode? >> >> IF it’s in time mode - what is the stop hooked to? >> >> IF it’s in frequent mode - what is the gate time set to ? >> >> Is the “standard input” the missing front panel input or is it the external >> reference input on the back panel? > > I was also considering the setup strange in this regard. Common switched in? > >> = >> >> I’m looking at the data on the link: >> >> http://www.ptp-images.com/affiche-directement-l-image-kccsz71c9a.html >> >> In both cases the slope is roughly 1/tau. Both plots end up in the 3 to 4 x >> 10^-13 range at 100 to 300 seconds. That’s suspiciously good performance >> for a rubidium or a GPSDO. Which is what makes me wonder about the setup. >> >> The Blue plot (1 pps?) ran for 18 minutes and has 1,114 points in it. The >> Pink plot ran for about 9 minutes and has a bit over 500 points in it. Both >> seem reasonable for a 1 pps to 10 MHz sort of setup. That may explain part >> of my confusion above. >> >> Again - I apologize if this all got explained in a post that went missing >> here. > > Not really. The plots looks to me like measurement setup
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Bob, On 01/18/2015 02:44 PM, Bob Camp wrote: Hi Ok, I didn’t think I’d seen the plots before. I agree that the plots look like “counter limited” data. That’s a fine explanation at the shorter Tau’s. I also agree that some sort of periodic “stuff” is getting into one of the signals and creating the ripple. Yes, that much is expected. What I’m wondering about (and what makes me question the setup) is the fact that the data is still “counter limited” at the mid to low parts in 10^-13 level at just a bit over 100 seconds. A telecom Rb is doing pretty well to be at 1x10^-12 at 100 seconds. Most GPSDO’s are doing well to be mid parts in 10^-12 at that tau. Simply put, the data continues to be counter limited to a pretty low point. In a setup where you feed both start and stop the same signal, the measurement noise of the counter setup will behave like that. Since the time-difference between the start and stop is very small, there is almost no lower-frequency noise included into the measurement, so the scaled variant of the noise will just keep going downwards. If you actuallly measure two sources, the initial slope will be that of the counter noise, but then pan out into the sum of the two sources. This is why I was considering that it was wired up as a instrument noise test. 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] question Alan deviation measured with Timelab and counters
Hi Effectively you only have one signal into the counter. It’s simply measuring the 1/tau measurement floor of the 5370. If your objective is to measure the ADEV of the measurement floor of the counter, that’s fine. Your plot looks a lot like the one half way down on: http://www.febo.com/pages/hp5370b/ It’s in the “Time Interval Measurement” section. ADEV of a standard is done by comparing two references against each other. With a 5370 typically one source starts a reading and the other source stops the reading. There are many different ways to set this up. A lot depends on the sources you have. One common way is: Set the counter to start / stop time interval mode. Let a PPS start the reading and a 10 MHz stop the reading. You will get data once a second. The time difference between start and stop will be the phase of the 10 MHz sampled once a second. The tricky part of this is getting all the buttons and knobs on the 5370 set to the right mode. It’s a fairly complex device. A quick way to debug the time mode on the counter: Feed the PPS to both inputs and have one trigger positive and the other trigger negative. You will read the pulse width (either high or low) of the PPS signal. The plot you will get will still be limited by the counter measurement floor at short time periods. Past 100 seconds or so you should be seeing the combined ADEV of your two sources. If the ADEV of each source is at the same level, then the sources are better than the reading by the square root of 2. If one source is much better than the other, the reading is essential the ADEV of the poorer source. Bob > On Jan 18, 2015, at 9:37 AM, Stéphane Rey wrote: > > Hello, > > First, please do apologize for the confusion answering in the bad email. > That's things I'm absolutely able to do when replying at 3 am ! Again, sorry > for that and thanks Magnus for having corrected this. > > Back to my setup : > > There is indeed nothing on the STOP input of the HP5370a. The standard 10 MHz > comes from the GPSDO HP-58503B and feeds the HP5370a Standard input. Its ADEV > is given on page 240 of that document : > http://www.leapsecond.com/museum/hp58503a/097-58503-13-iss-1.pdf We see that > the shape is starting at about 2E-12 at 1s, increase to 2E-11 at 100s before > decreasing again down to E-13 for above 10E3s... > > The setup #1 was using the Racal DANA Rb connected on the START input which > is specified at E-9 / E-10, given on page 16 of the manual : > http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Racal/9470-9479.pdf > The EXT input receives the 1PPS from the HP58503b. It apparently drives both > the START and the STOP of the acquisition (the two lights are blinking and > the time between two measurement is no longer adjustable from the front panel > RATE potentiometer and the period between two samples is 1.0s (detected by > Timelab). > > But yes, the ADEV plot sounds really strange as it goes incredibly low after > few seconds which is not consistent with the stability of the sources I'm > using which is why I felt something was wrong > > On Setup #2 I've only replaced the Racal Dana Rb with the GPSDO to test. I've > not made this design and not checked yet anything on it. Could these > oscillations be from power supply noise ? To be checked. But how can it > follow the ADEV plot of the Racal Dana Rb ? mmm Coincidence is not > something I like too much and I believe something is clearly wrong in my > measurement > > But what ??? > > On the Timelab setup screen before launching the acquisition I've left all > the parameters as it without touching them. I've just seize 10E6 in the > frequency field. > > Ah, by thay Magnus, for the downmixed test I've forgotten to change this > value, I will check on monday when back at the office. > > > Stephane > > > > > -Message d'origine----- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Bob Camp > Envoyé : dimanche 18 janvier 2015 14:44 > À : Discussion of precise time and frequency measurement > Objet : Re: [time-nuts] question Alan deviation measured with Timelab and > counters > > Hi > > Ok, I didn’t think I’d seen the plots before. > > I agree that the plots look like “counter limited” data. That’s a fine > explanation at the shorter Tau’s. I also agree that some sort of periodic > “stuff” is getting into one of the signals and creating the ripple. What I’m > wondering about (and what makes me question the setup) is the fact that the > data is still “counter limited” at the mid to low parts in 10^-13 level at > just a bit over 100 seconds. A telecom Rb is doing pretty well to be at > 1x10^-12 at 100 seconds. Most GPSD
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Bonjour Stéphane, On 01/18/2015 03:37 PM, Stéphane Rey wrote: Hello, First, please do apologize for the confusion answering in the bad email. That's things I'm absolutely able to do when replying at 3 am ! Again, sorry for that and thanks Magnus for having corrected this. Ah well, that's water under the bridge now. I only mentioned it for Bob's reference. Back to my setup : There is indeed nothing on the STOP input of the HP5370a. The standard 10 MHz comes from the GPSDO HP-58503B and feeds the HP5370a Standard input. Its ADEV is given on page 240 of that document : http://www.leapsecond.com/museum/hp58503a/097-58503-13-iss-1.pdf We see that the shape is starting at about 2E-12 at 1s, increase to 2E-11 at 100s before decreasing again down to E-13 for above 10E3s... The setup #1 was using the Racal DANA Rb connected on the START input which is specified at E-9 / E-10, given on page 16 of the manual : http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Racal/9470-9479.pdf The EXT input receives the 1PPS from the HP58503b. It apparently drives both the START and the STOP of the acquisition (the two lights are blinking and the time between two measurement is no longer adjustable from the front panel RATE potentiometer and the period between two samples is 1.0s (detected by Timelab). If you run the counter in frequency or period mode, you normally use the STOP input, which is then internally split to the START and STOP channels. If you run the counter in TI mode, then they are usually separate, but you can force them the same using the START COMMON switch. We tend to use the TI mode, with two basic setup: Stoopid simple: PPS to START and measured clock to STOP. This setup has the down-side that the jitter of the PPS (which can be much higher than that of the clock) can dominate, if so, the next setup is relevant: Standard setup: PPS to ARM/EXT input to trigger measurement. DUT to START channel and reference clock to STOP channel. Sometime the clocks is interchanged, sometimes it is important, somtimes not. Record the TI data. But yes, the ADEV plot sounds really strange as it goes incredibly low after few seconds which is not consistent with the stability of the sources I'm using which is why I felt something was wrong OK, you made what we call a instrument noise limit measurement. Then you do the same thing as a normal measure, but you have start and stop channels see the same signal split. It may be good to let the stop channel has a meter or two of additional coax to de-correlate the rising edges. This setup will let you measure the effect of white noise, slew-rate and counter resolution. It can be good for fault analysis and see if the setup gives reasonable noise or if you can improve it. Adjustment of the trigger points will select a point of optimal slew-rate (and sometimes avoid false-trigger noise) and thus finding the optimum trigger noise. Squaring up the signal may be a nice way to improve the setup. Anyway, such setup has the 1/tau plot behavior and that was what I saw. The fact that you kept going down was a clear clue that you where doing such a setup rather than doing a suitable delta. Now, try the two setups I proposed, letting the STOP channel being delayed with about 1 meter extra cable, and record the result. Do share for comments. Then, using the setup giving the lowest trace for measure your two other sources as DUT. On Setup #2 I've only replaced the Racal Dana Rb with the GPSDO to test. I've not made this design and not checked yet anything on it. Could these oscillations be from power supply noise ? To be checked. But how can it follow the ADEV plot of the Racal Dana Rb ? mmm Coincidence is not something I like too much and I believe something is clearly wrong in my measurement But what ??? Re-arranging the setup and it will be interesting to see both these setups. Then we can start making some comments on that result. On the Timelab setup screen before launching the acquisition I've left all the parameters as it without touching them. I've just seize 10E6 in the frequency field. Usually that's all that is needed. Ah, by thay Magnus, for the downmixed test I've forgotten to change this value, I will check on monday when back at the office. If you only have your TIM file with you back home, all you have to do is to press (e) to Edit the trace, as I recall it. I might have edited the file directly also. When doing that, I helped another time-nut at one time. Uncheck the "Use Input Frequency" and then input 10 MHz (or whatever) to "DUT Frequency". To actually make gains from a mixer-setup, you need to do more processing to filter and square up, but for the moment, it's just a nice lab-exercise. :) 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
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Thanks a lot Bob and Magnus for your very helpful comments. The HP5370a was indeed in TI mode. By the way what is the difference with +/-TI, the button just aside... But I guess I understand where I've missed something : I've tried to put the Rb on channel A and the DUT on channel B but result was always the same but I do understand now that there is indeed a switch to change from COMmon to SEParate and it was always on COM meaning I believe that channel B wasn't used. This explains a lot of things I did not understand. I'm sorry for these so basic issues that might have been solved if I had read carefully the HP5370a manual first. So possible conclusions until now are that I have actually measured the ADEV floor of the system rather than my DUT... which is already nice. The second conclusion from these oscillations seen with the GPSDO under test is that there is very likely in this GPSDO design a systemic noise added to the 10 MHz output (power supply, PCB coupling, ... I'll make further investigations on it later on). I will experiment all the suggestions you made and will come back. For information the 1PPS from the HP58503b has a positive pulse width that is only few us length. Now, when considering that the method is to compare the DUT to an other source, I assume then that the other source shall be at least 1 order of magnitude better than the DUT. Otherwise this will be impossible to distinguish who is the instability contributor between the source and DUT, right ? Then the second question is what kind of very stable source can be used to measure DUT which could be Rb or GPSDO which are already in the range of 10E-10 to 10E-12 < 100s ? Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Magnus Danielson Envoyé : dimanche 18 janvier 2015 16:47 À : time-nuts@febo.com Cc : mag...@rubidium.se Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Bonjour Stéphane, On 01/18/2015 03:37 PM, Stéphane Rey wrote: > Hello, > > First, please do apologize for the confusion answering in the bad email. > That's things I'm absolutely able to do when replying at 3 am ! Again, sorry > for that and thanks Magnus for having corrected this. Ah well, that's water under the bridge now. I only mentioned it for Bob's reference. > Back to my setup : > > There is indeed nothing on the STOP input of the HP5370a. The standard 10 MHz > comes from the GPSDO HP-58503B and feeds the HP5370a Standard input. Its ADEV > is given on page 240 of that document : > http://www.leapsecond.com/museum/hp58503a/097-58503-13-iss-1.pdf We see that > the shape is starting at about 2E-12 at 1s, increase to 2E-11 at 100s before > decreasing again down to E-13 for above 10E3s... > > The setup #1 was using the Racal DANA Rb connected on the START input > which is specified at E-9 / E-10, given on page 16 of the manual : > http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Racal/9470-9479.pdf > The EXT input receives the 1PPS from the HP58503b. It apparently drives both > the START and the STOP of the acquisition (the two lights are blinking and > the time between two measurement is no longer adjustable from the front panel > RATE potentiometer and the period between two samples is 1.0s (detected by > Timelab). If you run the counter in frequency or period mode, you normally use the STOP input, which is then internally split to the START and STOP channels. If you run the counter in TI mode, then they are usually separate, but you can force them the same using the START COMMON switch. We tend to use the TI mode, with two basic setup: Stoopid simple: PPS to START and measured clock to STOP. This setup has the down-side that the jitter of the PPS (which can be much higher than that of the clock) can dominate, if so, the next setup is relevant: Standard setup: PPS to ARM/EXT input to trigger measurement. DUT to START channel and reference clock to STOP channel. Sometime the clocks is interchanged, sometimes it is important, somtimes not. Record the TI data. > But yes, the ADEV plot sounds really strange as it goes incredibly low > after few seconds which is not consistent with the stability of the > sources I'm using which is why I felt something was wrong OK, you made what we call a instrument noise limit measurement. Then you do the same thing as a normal measure, but you have start and stop channels see the same signal split. It may be good to let the stop channel has a meter or two of additional coax to de-correlate the rising edges. This setup will let you measure the effect of white noise, slew-rate and counter resolution. It can be good for fault analysis and see if the setup gives reasonable noise or if you can improve it. Adjustment of the trigger points will select a poin
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi > On Jan 18, 2015, at 4:34 PM, Stéphane Rey wrote: > > Thanks a lot Bob and Magnus for your very helpful comments. > > The HP5370a was indeed in TI mode. By the way what is the difference with > +/-TI, the button just aside... > > But I guess I understand where I've missed something : I've tried to put the > Rb on channel A and the DUT on channel B but result was always the same but I > do understand now that there is indeed a switch to change from COMmon to > SEParate and it was always on COM meaning I believe that channel B wasn't > used. This explains a lot of things I did not understand. I'm sorry for these > so basic issues that might have been solved if I had read carefully the > HP5370a manual first. I’ve set that same switch the wrong way several times in the last year. It’s very frustrating …. > > So possible conclusions until now are that I have actually measured the ADEV > floor of the system rather than my DUT... which is already nice. The second > conclusion from these oscillations seen with the GPSDO under test is that > there is very likely in this GPSDO design a systemic noise added to the 10 > MHz output (power supply, PCB coupling, ... I'll make further investigations > on it later on). > > I will experiment all the suggestions you made and will come back. For > information the 1PPS from the HP58503b has a positive pulse width that is > only few us length. That sounds about right. > > Now, when considering that the method is to compare the DUT to an other > source, I assume then that the other source shall be at least 1 order of > magnitude better than the DUT. Otherwise this will be impossible to > distinguish who is the instability contributor between the source and DUT, > right ? If you can afford a source (or find a source) that’s 10X better, this would be the ideal solution. > > Then the second question is what kind of very stable source can be used to > measure DUT which could be Rb or GPSDO which are already in the range of > 10E-10 to 10E-12 < 100s ? Your GPSDO is probably around 2x10^-12 at 1 second. Your Rb probably gets to a floor of ~3x10^-13 at a few thousand seconds. An HP5071A with a high performance tube could be used to check the Rb at longer tau’s. A hydrogen maser could also be used. Buying a fully working version of either one will cost you the price of a nice new car (or a couple of new cars). A (rare) BVA OCXO could be used for shorter tau’s. A more practical (lower cost) approach is to compare multiple copies of the same device. With two Rb’s you can get a good guess at the worst they could each be doing. With three of them, you can make a pretty good guess about which one is best / which one is worst. The same approach also works with GPSDO’s. The catch on GPSDO’s is that the filters in them switch time constants automatically, so comparing them in an identical state can be tough. One way to get around problems with things like GPSDO’s is to have multiple standards and use them for different ranges of Tau. OCXO’s have good ADEV close in. Picking up a few and comparing them is pretty cheap. Rb’s have better ADEV at longer tau (> 300 sec). Comparing three or four is not as cheap as with OCXO’s but it’s not totally crazy. Running several GPSDO’s long enough to settle to their longest time constant is another way to go. Lots of choices. Lots of ways to clutter up the basement with TimeNut stuff… Bob > > > Stephane > > > -Message d'origine- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Magnus > Danielson > Envoyé : dimanche 18 janvier 2015 16:47 > À : time-nuts@febo.com > Cc : mag...@rubidium.se > Objet : Re: [time-nuts] question Alan deviation measured with Timelab and > counters > > Bonjour Stéphane, > > On 01/18/2015 03:37 PM, Stéphane Rey wrote: >> Hello, >> >> First, please do apologize for the confusion answering in the bad email. >> That's things I'm absolutely able to do when replying at 3 am ! Again, sorry >> for that and thanks Magnus for having corrected this. > > Ah well, that's water under the bridge now. I only mentioned it for Bob's > reference. > >> Back to my setup : >> >> There is indeed nothing on the STOP input of the HP5370a. The standard 10 >> MHz comes from the GPSDO HP-58503B and feeds the HP5370a Standard input. Its >> ADEV is given on page 240 of that document : >> http://www.leapsecond.com/museum/hp58503a/097-58503-13-iss-1.pdf We see that >> the shape is starting at about 2E-12 at 1s, increase to 2E-11 at 100s before >> decreasing again down to E-13 for above 10E3s... >> >> The setup #1 was using the Racal DANA Rb
Re: [time-nuts] question Alan deviation measured with Timelab and counters
> One way to get around problems with things like GPSDO’s is to have multiple > standards and use them for different ranges of Tau. OCXO’s have good ADEV > close in. Picking up a few and comparing them is pretty cheap. Rb’s have > better > ADEV at longer tau (> 300 sec). Comparing three or four is not as cheap as > with > OCXO’s but it’s not totally crazy. Running several GPSDO’s long enough to > settle to their longest time constant is another way to go. On that subject, I should probably mention that the current TimeLab beta (at http://www.miles.io/timelab/beta.htm ) has a realtime N-cornered hat display. I needed that feature to measure a bunch of OCXOs against a pair of Corby's trick 5065As, so I finally bit the bullet and added it. It is very cool, subject to the (many!) caveats with separated-variance measurements in general. This feature should work with measurements taken by counters as well, but I haven't tested that yet. Even worse, I won't have time in the immediate future to document it very extensively. Short version: measure three devices against each other, and then use the 'e'dit dialog to assign Source A and Source B labels to each of the resulting three plots. Go to the ADEV or other xDEV measurement view and hit ctrl-h to toggle the N-cornered hat display mode. The program will use your assigned source labels to separate the individual source variances. It won't work if you don't spell the source names consistently between each pair of measurements; if any of the measurements have different t0 intervals, bin counts, or trace history settings; if you don't read the mouseover help text for the Source A/B fields in the trace properties dialog; or if you don't read everything Bill Riley has written about 3-corner hat measurements. TimePod users can assign the source labels at acquisition time, via the new fields in the Advanced tab. Read the new mouseover text for the Stability and Ch0/Ch1/Ch2 fields _carefully_. Users of counters, or those who want to render existing files in 3-cornered hat mode, will need to add source labels in the 'e'dit dialog. Symmetricom/Microsemi 3120A users should email me offline if interested. I don't have a beta for the 3120A app yet, but it will be there eventually... -- john, KE5FX Miles Design LLC ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Bonsoir Stéphane, On 01/18/2015 10:34 PM, Stéphane Rey wrote: Thanks a lot Bob and Magnus for your very helpful comments. The HP5370a was indeed in TI mode. By the way what is the difference with +/-TI, the button just aside... But I guess I understand where I've missed something : I've tried to put the Rb on channel A and the DUT on channel B but result was always the same but I do understand now that there is indeed a switch to change from COMmon to SEParate and it was always on COM meaning I believe that channel B wasn't used. This explains a lot of things I did not understand. I'm sorry for these so basic issues that might have been solved if I had read carefully the HP5370a manual first. Good. This confirmation makes sense to be and Bob, now we can relax as the mystery is solved. So possible conclusions until now are that I have actually measured the ADEV floor of the system rather than my DUT... which is already nice. The second conclusion from these oscillations seen with the GPSDO under test is that there is very likely in this GPSDO design a systemic noise added to the 10 MHz output (power supply, PCB coupling, ... I'll make further investigations on it later on). It's a great opportunity to learn the tools, and once you have the tools, you can see if you can't improve things. I will experiment all the suggestions you made and will come back. For information the 1PPS from the HP58503b has a positive pulse width that is only few us length. This only makes it hard to view on a scope, but long enough to reliably trigger your counter and scope. Now, when considering that the method is to compare the DUT to an other source, I assume then that the other source shall be at least 1 order of magnitude better than the DUT. Otherwise this will be impossible to distinguish who is the instability contributor between the source and DUT, right ? For a simple setup, yes. But then we are the time-nuts, we have ways of handling these things. :) Let's get you started with the basic measurement, it will be a good start. Then the second question is what kind of very stable source can be used to measure DUT which could be Rb or GPSDO which are already in the range of 10E-10 to 10E-12 < 100s ? Time-nuts tend to spend their time and money getting even more stable clocks and tools. If you have the right tool, you can measure near and *under* the noise-level of your reference, but not without running into issues. One such trick is called cross-correlation, while another is to use three-corner hat techniques. 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] question Alan deviation measured with Timelab and counters
Bonsoir Magnus (Are you in Sweeden ?) Being able to measure high stability and low phase noise is definitely a need for me as I'm trying to design low noise synthesizers and I'm already reaching the limits of my current tools for phase noise and I can't afford an E5052 for my own. At work I've one but I will probably not stay after august. And anyway I need such tools in my lab at home... As low-noise and stable synthetizers depends on the standard used, I need as well to measure them as well... Let's start with this simple experiments and once I will understand the ins and outs I will try to improve. I know techniques of cross-correlations and you've already talked about DMTD that for sure I will have to come to... Good night Stephane -Message d'origine- De : Magnus Danielson [mailto:mag...@rubidium.se] Envoyé : dimanche 18 janvier 2015 22:46 À : Stéphane Rey; 'Discussion of precise time and frequency measurement' Cc : mag...@rubidium.se Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Bonsoir Stéphane, On 01/18/2015 10:34 PM, Stéphane Rey wrote: > Thanks a lot Bob and Magnus for your very helpful comments. > > The HP5370a was indeed in TI mode. By the way what is the difference with > +/-TI, the button just aside... > > But I guess I understand where I've missed something : I've tried to put the > Rb on channel A and the DUT on channel B but result was always the same but I > do understand now that there is indeed a switch to change from COMmon to > SEParate and it was always on COM meaning I believe that channel B wasn't > used. This explains a lot of things I did not understand. I'm sorry for these > so basic issues that might have been solved if I had read carefully the > HP5370a manual first. Good. This confirmation makes sense to be and Bob, now we can relax as the mystery is solved. > So possible conclusions until now are that I have actually measured the ADEV > floor of the system rather than my DUT... which is already nice. The second > conclusion from these oscillations seen with the GPSDO under test is that > there is very likely in this GPSDO design a systemic noise added to the 10 > MHz output (power supply, PCB coupling, ... I'll make further investigations > on it later on). It's a great opportunity to learn the tools, and once you have the tools, you can see if you can't improve things. > I will experiment all the suggestions you made and will come back. For > information the 1PPS from the HP58503b has a positive pulse width that is > only few us length. This only makes it hard to view on a scope, but long enough to reliably trigger your counter and scope. > Now, when considering that the method is to compare the DUT to an other > source, I assume then that the other source shall be at least 1 order of > magnitude better than the DUT. Otherwise this will be impossible to > distinguish who is the instability contributor between the source and DUT, > right ? For a simple setup, yes. But then we are the time-nuts, we have ways of handling these things. :) Let's get you started with the basic measurement, it will be a good start. > Then the second question is what kind of very stable source can be used to > measure DUT which could be Rb or GPSDO which are already in the range of > 10E-10 to 10E-12 < 100s ? Time-nuts tend to spend their time and money getting even more stable clocks and tools. If you have the right tool, you can measure near and *under* the noise-level of your reference, but not without running into issues. One such trick is called cross-correlation, while another is to use three-corner hat techniques. Cheers, Magnus --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.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] question Alan deviation measured with Timelab and counters
Hi > On Jan 18, 2015, at 5:12 PM, Stéphane Rey wrote: > > Bonsoir Magnus (Are you in Sweeden ?) > > Being able to measure high stability and low phase noise is definitely a need > for me as I'm trying to design low noise synthesizers and I'm already > reaching the limits of my current tools for phase noise and I can't afford an > E5052 for my own. At work I've one but I will probably not stay after august. > And anyway I need such tools in my lab at home… If you have tools at work, the best possible thing to do is to get some oscillators / standards characterized. If you *know* what this or that oscillator is doing in terms of ADEV or phase noise at this Tau or frequency offset, it’s much easier to figure a lot of this out. The most basic way to do phase noise in the basement is with a single mixer setup running into some sort of audio FFT device. A sound card can be used or an audio spectrum analyzer. Parts are < $100 to get one setup once you can do the audio measurements. For ADEV, a DMTD or it’s cousin, the single mixer is the easy way to go. The single mixer does not get a lot of discussion these days. It is much easier to set up than a DMTD. It does require an offset oscillator. Once you have a single mixer phase noise setup, you are about half way to a single mixer ADEV setup. Cost for one is < $100 in parts. You already have a counter to collect the data out of it. In both cases you are running a comparison device. Having a characterized OCXO to compare to is a really nice thing. Bob > As low-noise and stable synthetizers depends on the standard used, I need as > well to measure them as well... > > Let's start with this simple experiments and once I will understand the ins > and outs I will try to improve. I know techniques of cross-correlations and > you've already talked about DMTD that for sure I will have to come to... > > Good night > Stephane > > -Message d'origine- > De : Magnus Danielson [mailto:mag...@rubidium.se] > Envoyé : dimanche 18 janvier 2015 22:46 > À : Stéphane Rey; 'Discussion of precise time and frequency measurement' > Cc : mag...@rubidium.se > Objet : Re: [time-nuts] question Alan deviation measured with Timelab and > counters > > Bonsoir Stéphane, > > On 01/18/2015 10:34 PM, Stéphane Rey wrote: >> Thanks a lot Bob and Magnus for your very helpful comments. >> >> The HP5370a was indeed in TI mode. By the way what is the difference with >> +/-TI, the button just aside... >> >> But I guess I understand where I've missed something : I've tried to put the >> Rb on channel A and the DUT on channel B but result was always the same but >> I do understand now that there is indeed a switch to change from COMmon to >> SEParate and it was always on COM meaning I believe that channel B wasn't >> used. This explains a lot of things I did not understand. I'm sorry for >> these so basic issues that might have been solved if I had read carefully >> the HP5370a manual first. > > Good. This confirmation makes sense to be and Bob, now we can relax as the > mystery is solved. > >> So possible conclusions until now are that I have actually measured the ADEV >> floor of the system rather than my DUT... which is already nice. The second >> conclusion from these oscillations seen with the GPSDO under test is that >> there is very likely in this GPSDO design a systemic noise added to the 10 >> MHz output (power supply, PCB coupling, ... I'll make further investigations >> on it later on). > > It's a great opportunity to learn the tools, and once you have the tools, you > can see if you can't improve things. > >> I will experiment all the suggestions you made and will come back. For >> information the 1PPS from the HP58503b has a positive pulse width that is >> only few us length. > > This only makes it hard to view on a scope, but long enough to reliably > trigger your counter and scope. > >> Now, when considering that the method is to compare the DUT to an other >> source, I assume then that the other source shall be at least 1 order of >> magnitude better than the DUT. Otherwise this will be impossible to >> distinguish who is the instability contributor between the source and DUT, >> right ? > > For a simple setup, yes. But then we are the time-nuts, we have ways of > handling these things. :) Let's get you started with the basic measurement, > it will be a good start. > >> Then the second question is what kind of very stable source can be used to >> measure DUT which could be Rb or GPSDO which are already i
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Actually I'm working in the RF department of a big lab, designing RF electronics mainly in microwaves range. I'm luckilly having some tools around to play with and a lot of components like mixers/amplifiers/couplers/splitters/attenuators, ... almost whatever the frequency is up to several tens of GHz. At home since the last 20 years I could as well get nice instruments. The next two measuring tools really missing and for which I'm limited are the phase noise and stability measurement and possibly a good standard. My Effratom FRK Rb is old and probably not the best from a phase noise and stability point of view but until now has never been characterized. Otherwise I've almost everything I need up to 40 GHz I guess. I'm doing further measurement right now which sounds much much more consistent. I will share tonight. Cheers Stephane On Mon, 19 Jan 2015 08:59:58 -0500, Bob Camp wrote: Hi On Jan 18, 2015, at 5:12 PM, Stéphane Rey wrote: Bonsoir Magnus (Are you in Sweeden ?) Being able to measure high stability and low phase noise is definitely a need for me as I'm trying to design low noise synthesizers and I'm already reaching the limits of my current tools for phase noise and I can't afford an E5052 for my own. At work I've one but I will probably not stay after august. And anyway I need such tools in my lab at home… If you have tools at work, the best possible thing to do is to get some oscillators / standards characterized. If you *know* what this or that oscillator is doing in terms of ADEV or phase noise at this Tau or frequency offset, it’s much easier to figure a lot of this out. The most basic way to do phase noise in the basement is with a single mixer setup running into some sort of audio FFT device. A sound card can be used or an audio spectrum analyzer. Parts are < $100 to get one setup once you can do the audio measurements. For ADEV, a DMTD or it’s cousin, the single mixer is the easy way to go. The single mixer does not get a lot of discussion these days. It is much easier to set up than a DMTD. It does require an offset oscillator. Once you have a single mixer phase noise setup, you are about half way to a single mixer ADEV setup. Cost for one is < $100 in parts. You already have a counter to collect the data out of it. In both cases you are running a comparison device. Having a characterized OCXO to compare to is a really nice thing. Bob As low-noise and stable synthetizers depends on the standard used, I need as well to measure them as well... Let's start with this simple experiments and once I will understand the ins and outs I will try to improve. I know techniques of cross-correlations and you've already talked about DMTD that for sure I will have to come to... Good night Stephane -Message d'origine- De : Magnus Danielson [mailto:mag...@rubidium.se] Envoyé : dimanche 18 janvier 2015 22:46 À : Stéphane Rey; 'Discussion of precise time and frequency measurement' Cc : mag...@rubidium.se Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Bonsoir Stéphane, On 01/18/2015 10:34 PM, Stéphane Rey wrote: Thanks a lot Bob and Magnus for your very helpful comments. The HP5370a was indeed in TI mode. By the way what is the difference with +/-TI, the button just aside... But I guess I understand where I've missed something : I've tried to put the Rb on channel A and the DUT on channel B but result was always the same but I do understand now that there is indeed a switch to change from COMmon to SEParate and it was always on COM meaning I believe that channel B wasn't used. This explains a lot of things I did not understand. I'm sorry for these so basic issues that might have been solved if I had read carefully the HP5370a manual first. Good. This confirmation makes sense to be and Bob, now we can relax as the mystery is solved. So possible conclusions until now are that I have actually measured the ADEV floor of the system rather than my DUT... which is already nice. The second conclusion from these oscillations seen with the GPSDO under test is that there is very likely in this GPSDO design a systemic noise added to the 10 MHz output (power supply, PCB coupling, ... I'll make further investigations on it later on). It's a great opportunity to learn the tools, and once you have the tools, you can see if you can't improve things. I will experiment all the suggestions you made and will come back. For information the 1PPS from the HP58503b has a positive pulse width that is only few us length. This only makes it hard to view on a scope, but long enough to reliably trigger your counter and scope. Now, when considering that the method is to compare the DUT to an other source, I assume then that the other source shall be at least 1 order of magnitude better than th
Re: [time-nuts] question Alan deviation measured with Timelab and counters
John, Your new three-corner hat feature is really cool. Already tried it, even if I did not spend quality time on setting source labels correctly. Whenever I have a setup capable of running it, I will test-spin it again. I suspect that when you make separate measurements, the noise of the individual sources does not match up perfectly which may reduce quality, especially systematics like hum. Cheers, Magnus On 01/19/2015 03:23 AM, John Miles wrote: One way to get around problems with things like GPSDO’s is to have multiple standards and use them for different ranges of Tau. OCXO’s have good ADEV close in. Picking up a few and comparing them is pretty cheap. Rb’s have better ADEV at longer tau (> 300 sec). Comparing three or four is not as cheap as with OCXO’s but it’s not totally crazy. Running several GPSDO’s long enough to settle to their longest time constant is another way to go. On that subject, I should probably mention that the current TimeLab beta (at http://www.miles.io/timelab/beta.htm ) has a realtime N-cornered hat display. I needed that feature to measure a bunch of OCXOs against a pair of Corby's trick 5065As, so I finally bit the bullet and added it. It is very cool, subject to the (many!) caveats with separated-variance measurements in general. This feature should work with measurements taken by counters as well, but I haven't tested that yet. Even worse, I won't have time in the immediate future to document it very extensively. Short version: measure three devices against each other, and then use the 'e'dit dialog to assign Source A and Source B labels to each of the resulting three plots. Go to the ADEV or other xDEV measurement view and hit ctrl-h to toggle the N-cornered hat display mode. The program will use your assigned source labels to separate the individual source variances. It won't work if you don't spell the source names consistently between each pair of measurements; if any of the measurements have different t0 intervals, bin counts, or trace history settings; if you don't read the mouseover help text for the Source A/B fields in the trace properties dialog; or if you don't read everything Bill Riley has written about 3-corner hat measurements. TimePod users can assign the source labels at acquisition time, via the new fields in the Advanced tab. Read the new mouseover text for the Stability and Ch0/Ch1/Ch2 fields _carefully_. Users of counters, or those who want to render existing files in 3-cornered hat mode, will need to add source labels in the 'e'dit dialog. Symmetricom/Microsemi 3120A users should email me offline if interested. I don't have a beta for the 3120A app yet, but it will be there eventually... -- john, KE5FX Miles Design LLC ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
> John, > > Your new three-corner hat feature is really cool. Already tried it, even > if I did not spend quality time on setting source labels correctly. > > Whenever I have a setup capable of running it, I will test-spin it again. > > I suspect that when you make separate measurements, the noise of the > individual sources does not match up perfectly which may reduce quality, > especially systematics like hum. True, and Bill Riley's documents emphasize that simultaneous measurements are best. That's easy to do with the SMA jacks on the TimePod but more of a challenge with counters, unless you buy several of them and run them from separate GPIB adapters or serial dongles. IMO, if you can keep correlation, tonal artifacts, and other systematic errors under control and make long-enough measurements to generate a lot of data points in the tau range of interest, then the technique should work well with non-simultaneous acquisitions. At the end of the day it's no better or worse than the quality of the individual ADEV measurements. You have to be wary of measurements at taus where any of the traces have large error bars or are wandering up and down over time, or where some of the traces are suspiciously close to each other. -- john, KE5FX Miles Design LLC ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
On Mon, 19 Jan 2015 08:59:58 -0500 Bob Camp wrote: > The most basic way to do phase noise in the basement is with a single mixer > setup running into some sort of audio FFT device. A sound card can be used or > an audio spectrum analyzer. Parts are < $100 to get one setup once you can do > the audio measurements. 2-3 years ago, i got a presentation of an italian amateur radio on how to measure phase noise of osciallators using this technique in quite detail. Only draw back is that the presentation is in italian. If someone wants this presentation, please contact me off list. Attila Kinali -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hi, Following the tests results in the previous email, today I've performed additional measurements showing that the repeatability of the GPSDO DUT is not great but is coming from the design. I've tested several over sources and repeatability is correct. I can already make some measurement. Good ! Now I'd like to improve. First I'm going to implement a squarer and then I will work on the DMTD... I'm thinking to make a setup on the table, and possibly make a small PCB then. Any comment for the tests results of yesterday here under ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : lundi 19 janvier 2015 22:32 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi Here are the results of today's experiments. plots and TIM files attached to this email. Setup #1 : dark blue I've done again the floor measurement with same conditions : HP58503 for 10 MHz Standard, 1PPS for the EXT gating and the Rb on channel A. Same result (hopefully) Setup #2 : Pink Then I've made what Magnus has suggested, i.e. using the 1 PPS on Channel A, the Rb on channel B and internal gating The ADEV has increased by more than 1 order of magnitude. I guess this confirms the 1PPS stability is lower than the 10 MHz Setup #3-6 : Dark Green, Red, Light blue and Dark yellow. I've measured several times the GPSDO DUT with SEParate inputs. 1PPS on EXT, Rb on channel A and DUT on channel B. This gives 4 different plots... When starting the measurement the plots starts directly at different values... Mmmm very strange. Is it coming for the setup of the GPSDO ? To be investigated further with other sources. This is the plan for tomorrow. However the overal shape of the plot sounds relevant to me. Setup #7-8 + #9 not showed here I've tested the suggested splitted same signal on both inputs with 1m coax for channel B. I've discovered that when swaping the GPSDO on the standard input and the Rb on the channel A I have a slight difference. In order to confirm I've made two time each measurement and this confirms that having the Rb on channel A and GPSDO on the standard input gives the lowest ADEV. The setup #9 which is the same than the light green gives the superimposed plot on that one... So what does it mean ? One of the two sources is better than the other, but which one ? Some other comments : - Swaping signals between channel A and B gives the same ADEV (setup #4 and 5, light blue and red) - On some measurement on the GPSDO DUT, (not displayed here), I could see during the measurement suddenly an increase of one order of magnitude. The HP5370A do not show any difference (the time interval value continues to move with a beat but visually impossible to quantify if the value between two values has increased. No explanation for that. I'll redo the test with some other sources to check if it comes for the measurement system or the GPSDO DUT In conclusion, 1. swaping the Rb and HP58503 doesn't give the same result. The GPSDO has standard seems the best (or the Rb measured) 2. the measurement on the GPSDO DUT gives different results with nearly one order of magnitude difference but shape is still the same. 3. the 1PPS must be connected on the EXT gating input What do you think ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de steph.rey Envoyé : lundi 19 janvier 2015 16:44 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Actually I'm working in the RF department of a big lab, designing RF electronics mainly in microwaves range. I'm luckilly having some tools around to play with and a lot of components like mixers/amplifiers/couplers/splitters/attenuators, ... almost whatever the frequency is up to several tens of GHz. At home since the last 20 years I could as well get nice instruments. The next two measuring tools really missing and for which I'm limited are the phase noise and stability measurement and possibly a good standard. My Effratom FRK Rb is old and probably not the best from a phase noise and stability point of view but until now has never been characterized. Otherwise I've almost everything I need up to 40 GHz I guess. I'm doing further measurement right now which sounds much much more consistent. I will share tonight. Cheers Stephane On Mon, 19 Jan 2015 08:59:58 -0500, Bob Camp wrote: > Hi > >> On Jan 18, 2015, at 5:12 PM, Stéphane Rey >> wrote: >> >> Bonsoir Magnus (Are you in Sweeden ?) >> >> Being able to measure high stability and low phase noise is >> definit
Re: [time-nuts] question Alan deviation measured with Timelab and counters
John, On 01/20/2015 01:55 AM, John Miles wrote: John, Your new three-corner hat feature is really cool. Already tried it, even if I did not spend quality time on setting source labels correctly. Whenever I have a setup capable of running it, I will test-spin it again. I suspect that when you make separate measurements, the noise of the individual sources does not match up perfectly which may reduce quality, especially systematics like hum. True, and Bill Riley's documents emphasize that simultaneous measurements are best. That's easy to do with the SMA jacks on the TimePod but more of a challenge with counters, unless you buy several of them and run them from separate GPIB adapters or serial dongles. IMO, if you can keep correlation, tonal artifacts, and other systematic errors under control and make long-enough measurements to generate a lot of data points in the tau range of interest, then the technique should work well with non-simultaneous acquisitions. At the end of the day it's no better or worse than the quality of the individual ADEV measurements. You have to be wary of measurements at taus where any of the traces have large error bars or are wandering up and down over time, or where some of the traces are suspiciously close to each other. Agree. For oscillators, they should have been turned on long enough such that any drift is negligible. Alternatively you process out the quadratic trend out of it. The later should be accompanied by some quality measure of how much remaining systematics there is (see Jim Barnes PTTI paper on Drift Estimators). Also, identifying systematic noises (such as in phase-noise plot) and notching them out before processing should improve the quality of the 3/N-cornered hat processing, since it really is for solving power-sums of random noise, but not for systematic noise. Until such processing of each measurement is in place, we have to make as clean measurement runs as possible, which is the basic quality of the ADEV. 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] question Alan deviation measured with Timelab and counters
Hi guys. After several experiments I could discover that the "bad" ADEV from the two GPSDO DUT are due to GPS lock losses. This is probably because the antenna is outside the windows but half the sky is hidden. We can see the on the frequency plot the sharp change of 0.5Hz and the locking. Good point. I'm now trying to evaluate various architectures of 2-channels squarers and a DMDT. For that I'm designing a PCB with 4 squarers : simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the transistor based differential amplifier from Winzel and the one from Charles. I will add two balanced mixers (minicircuits), IF filters and amplifiers. Does anyone has an idea of what I could add for this evaluation ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : mardi 20 janvier 2015 23:15 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi, Following the tests results in the previous email, today I've performed additional measurements showing that the repeatability of the GPSDO DUT is not great but is coming from the design. I've tested several over sources and repeatability is correct. I can already make some measurement. Good ! Now I'd like to improve. First I'm going to implement a squarer and then I will work on the DMTD... I'm thinking to make a setup on the table, and possibly make a small PCB then. Any comment for the tests results of yesterday here under ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : lundi 19 janvier 2015 22:32 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi Here are the results of today's experiments. plots and TIM files attached to this email. Setup #1 : dark blue I've done again the floor measurement with same conditions : HP58503 for 10 MHz Standard, 1PPS for the EXT gating and the Rb on channel A. Same result (hopefully) Setup #2 : Pink Then I've made what Magnus has suggested, i.e. using the 1 PPS on Channel A, the Rb on channel B and internal gating The ADEV has increased by more than 1 order of magnitude. I guess this confirms the 1PPS stability is lower than the 10 MHz Setup #3-6 : Dark Green, Red, Light blue and Dark yellow. I've measured several times the GPSDO DUT with SEParate inputs. 1PPS on EXT, Rb on channel A and DUT on channel B. This gives 4 different plots... When starting the measurement the plots starts directly at different values... Mmmm very strange. Is it coming for the setup of the GPSDO ? To be investigated further with other sources. This is the plan for tomorrow. However the overal shape of the plot sounds relevant to me. Setup #7-8 + #9 not showed here I've tested the suggested splitted same signal on both inputs with 1m coax for channel B. I've discovered that when swaping the GPSDO on the standard input and the Rb on the channel A I have a slight difference. In order to confirm I've made two time each measurement and this confirms that having the Rb on channel A and GPSDO on the standard input gives the lowest ADEV. The setup #9 which is the same than the light green gives the superimposed plot on that one... So what does it mean ? One of the two sources is better than the other, but which one ? Some other comments : - Swaping signals between channel A and B gives the same ADEV (setup #4 and 5, light blue and red) - On some measurement on the GPSDO DUT, (not displayed here), I could see during the measurement suddenly an increase of one order of magnitude. The HP5370A do not show any difference (the time interval value continues to move with a beat but visually impossible to quantify if the value between two values has increased. No explanation for that. I'll redo the test with some other sources to check if it comes for the measurement system or the GPSDO DUT In conclusion, 1. swaping the Rb and HP58503 doesn't give the same result. The GPSDO has standard seems the best (or the Rb measured) 2. the measurement on the GPSDO DUT gives different results with nearly one order of magnitude difference but shape is still the same. 3. the 1PPS must be connected on the EXT gating input What do you think ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de steph.rey Envoyé : lundi 19 janvier 2015 16:44 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Actually I'm working in the RF department of a big lab, designing RF electronics mainly in microwav
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi, Just a stupid question on Timelab. Why do I have the plot with 1/4 for the time actually used for the measurement ? I can see that the plot is updated every 4 samples but the scale is not relevant. The sample interval is correctly set (1s) Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : mardi 20 janvier 2015 23:15 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi, Following the tests results in the previous email, today I've performed additional measurements showing that the repeatability of the GPSDO DUT is not great but is coming from the design. I've tested several over sources and repeatability is correct. I can already make some measurement. Good ! Now I'd like to improve. First I'm going to implement a squarer and then I will work on the DMTD... I'm thinking to make a setup on the table, and possibly make a small PCB then. Any comment for the tests results of yesterday here under ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : lundi 19 janvier 2015 22:32 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi Here are the results of today's experiments. plots and TIM files attached to this email. Setup #1 : dark blue I've done again the floor measurement with same conditions : HP58503 for 10 MHz Standard, 1PPS for the EXT gating and the Rb on channel A. Same result (hopefully) Setup #2 : Pink Then I've made what Magnus has suggested, i.e. using the 1 PPS on Channel A, the Rb on channel B and internal gating The ADEV has increased by more than 1 order of magnitude. I guess this confirms the 1PPS stability is lower than the 10 MHz Setup #3-6 : Dark Green, Red, Light blue and Dark yellow. I've measured several times the GPSDO DUT with SEParate inputs. 1PPS on EXT, Rb on channel A and DUT on channel B. This gives 4 different plots... When starting the measurement the plots starts directly at different values... Mmmm very strange. Is it coming for the setup of the GPSDO ? To be investigated further with other sources. This is the plan for tomorrow. However the overal shape of the plot sounds relevant to me. Setup #7-8 + #9 not showed here I've tested the suggested splitted same signal on both inputs with 1m coax for channel B. I've discovered that when swaping the GPSDO on the standard input and the Rb on the channel A I have a slight difference. In order to confirm I've made two time each measurement and this confirms that having the Rb on channel A and GPSDO on the standard input gives the lowest ADEV. The setup #9 which is the same than the light green gives the superimposed plot on that one... So what does it mean ? One of the two sources is better than the other, but which one ? Some other comments : - Swaping signals between channel A and B gives the same ADEV (setup #4 and 5, light blue and red) - On some measurement on the GPSDO DUT, (not displayed here), I could see during the measurement suddenly an increase of one order of magnitude. The HP5370A do not show any difference (the time interval value continues to move with a beat but visually impossible to quantify if the value between two values has increased. No explanation for that. I'll redo the test with some other sources to check if it comes for the measurement system or the GPSDO DUT In conclusion, 1. swaping the Rb and HP58503 doesn't give the same result. The GPSDO has standard seems the best (or the Rb measured) 2. the measurement on the GPSDO DUT gives different results with nearly one order of magnitude difference but shape is still the same. 3. the 1PPS must be connected on the EXT gating input What do you think ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de steph.rey Envoyé : lundi 19 janvier 2015 16:44 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Actually I'm working in the RF department of a big lab, designing RF electronics mainly in microwaves range. I'm luckilly having some tools around to play with and a lot of components like mixers/amplifiers/couplers/splitters/attenuators, ... almost whatever the frequency is up to several tens of GHz. At home since the last 20 years I could as well get nice instruments. The next two measuring tools really missing and for which I'm limited are the phase noise and stability measurement and possibly a good standard. My Effratom FRK Rb is old and probably not the best from a p
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi If you go back in the archives and look for the discussions on “Collins hard limiter” they will lead you to some other areas to consider when trying to square low frequency sine wave signals. The quick summary is that you need a series of bandwidth limited limiter stages ahead of what ever you use as a squaring circuit. There are a number of ways to make these stages, each with their benefits and drawbacks. The limiting process has a much larger impact on the results than the choice of squaring circuit. Bob > On Jan 24, 2015, at 3:07 PM, Stéphane Rey wrote: > > Hi guys. > > After several experiments I could discover that the "bad" ADEV from the two > GPSDO DUT are due to GPS lock losses. This is probably because the antenna is > outside the windows but half the sky is hidden. We can see the on the > frequency plot the sharp change of 0.5Hz and the locking. Good point. > > I'm now trying to evaluate various architectures of 2-channels squarers and a > DMDT. For that I'm designing a PCB with 4 squarers : simple 74ac04 gate > biased at VCC/2, a LT1016 comparator, the transistor based differential > amplifier from Winzel and the one from Charles. I will add two balanced > mixers (minicircuits), IF filters and amplifiers. > Does anyone has an idea of what I could add for this evaluation ? > > Cheers > Stephane > > > -Message d'origine- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey > Envoyé : mardi 20 janvier 2015 23:15 > À : 'Discussion of precise time and frequency measurement' > Objet : Re: [time-nuts] question Alan deviation measured with Timelab and > counters > > Hi, > > Following the tests results in the previous email, today I've performed > additional measurements showing that the repeatability of the GPSDO DUT is > not great but is coming from the design. I've tested several over sources and > repeatability is correct. > > I can already make some measurement. Good ! > > Now I'd like to improve. First I'm going to implement a squarer and then I > will work on the DMTD... I'm thinking to make a setup on the table, and > possibly make a small PCB then. > > Any comment for the tests results of yesterday here under ? > > Cheers > Stephane > > > -Message d'origine----- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey > Envoyé : lundi 19 janvier 2015 22:32 À : 'Discussion of precise time and > frequency measurement' > Objet : Re: [time-nuts] question Alan deviation measured with Timelab and > counters > > Hi > > Here are the results of today's experiments. plots and TIM files attached to > this email. > > > Setup #1 : dark blue > I've done again the floor measurement with same conditions : HP58503 for 10 > MHz Standard, 1PPS for the EXT gating and the Rb on channel A. Same result > (hopefully) > > Setup #2 : Pink > Then I've made what Magnus has suggested, i.e. using the 1 PPS on Channel A, > the Rb on channel B and internal gating The ADEV has increased by more > than 1 order of magnitude. I guess this confirms the 1PPS stability is lower > than the 10 MHz > > Setup #3-6 : Dark Green, Red, Light blue and Dark yellow. > I've measured several times the GPSDO DUT with SEParate inputs. 1PPS on EXT, > Rb on channel A and DUT on channel B. This gives 4 different plots... When > starting the measurement the plots starts directly at different values... > Mmmm very strange. Is it coming for the setup of the GPSDO ? To be > investigated further with other sources. This is the plan for tomorrow. > However the overal shape of the plot sounds relevant to me. > > Setup #7-8 + #9 not showed here > I've tested the suggested splitted same signal on both inputs with 1m coax > for channel B. I've discovered that when swaping the GPSDO on the standard > input and the Rb on the channel A I have a slight difference. In order to > confirm I've made two time each measurement and this confirms that having the > Rb on channel A and GPSDO on the standard input gives the lowest ADEV. The > setup #9 which is the same than the light green gives the superimposed plot > on that one... So what does it mean ? One of the two sources is better than > the other, but which one ? > > > Some other comments : > - Swaping signals between channel A and B gives the same ADEV (setup #4 and > 5, light blue and red) > - On some measurement on the GPSDO DUT, (not displayed here), I could see > during the measurement suddenly an increase of one order of magnitude. The > HP5370A do not show any difference (t
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Hi Stéphane, have you read W. Riley's paper on a DMTD system? http://www.wriley.com/A%20Small%20DMTD%20System.pdf Cheers, Adrian Stéphane Rey schrieb: > Hi guys. > > After several experiments I could discover that the "bad" ADEV from the two > GPSDO DUT are due to GPS lock losses. This is probably because the antenna is > outside the windows but half the sky is hidden. We can see the on the > frequency plot the sharp change of 0.5Hz and the locking. Good point. > > I'm now trying to evaluate various architectures of 2-channels squarers and a > DMDT. For that I'm designing a PCB with 4 squarers : simple 74ac04 gate > biased at VCC/2, a LT1016 comparator, the transistor based differential > amplifier from Winzel and the one from Charles. I will add two balanced > mixers (minicircuits), IF filters and amplifiers. > Does anyone has an idea of what I could add for this evaluation ? > > Cheers > Stephane > snip ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Bob et Stéphane, The "Collins hard limiter" is a fancy squarer circuit. In short, the Collins hard limiter is what I hinted about in my earlier message. Rather than hitting the hard limiter of the comparator directly, with the slew-rate issue I gave. Collins observed that if you provide a linear amplifier (with limiting), you can increase the slew-rate of the signal. At the same time you will add noise from the amplifier. Now, to limit the noise going forward, you bandwidth limit the amplifier, which is fine as long as it supports the outgoing slew-rate of that amp. This helps you a bit and many times a single linear stage suffice to make the trigger noise sufficiently low, but then you can cascade these, and for many research DMTD setups has some hand-hacked variant. Collins then made a systematic approach to optimize the bandwidth and gain for a number of stages. Collins did however not think about amplifiers with different noise-contributions, but Bruce Griffiths have then generalized it further. This makes sense, since for higher slew-rates, you need faster amps, but their noise-contribution will be fairly low considering the high slew-rate coming into it. The original Collins article is not freely available. Therefore I recommend you to dip your nose into Bruce pages: http://www.ko4bb.com/~bruce/ Zero Crossing Detectors and Collins: http://www.ko4bb.com/~bruce/ZeroCrossingDetectors.html paper: http://www.ko4bb.com/~bruce/GeneralisedCollinsHardLimiterPaperV3B.pdf So, "squaring up" with 74AC04 or comparator and similar stuff might be an interesting exercise, but you end up doing the same exercise as going straight into the counters comparator. Try to amplify yourself out of the situation, try different gain settings. Make the habit of measuring the slew-rate (at the trigger point). It's interesting to note that not many counter support the feature of measuring slew-rate directly, the closest one usually get is to get the rise-time, but it needs scaling with the difference of the start and stop trigger-points, and you want them not at 10%-90% but closer. Cheers, Magnus On 01/24/2015 11:36 PM, Bob Camp wrote: Hi If you go back in the archives and look for the discussions on “Collins hard limiter” they will lead you to some other areas to consider when trying to square low frequency sine wave signals. The quick summary is that you need a series of bandwidth limited limiter stages ahead of what ever you use as a squaring circuit. There are a number of ways to make these stages, each with their benefits and drawbacks. The limiting process has a much larger impact on the results than the choice of squaring circuit. Bob On Jan 24, 2015, at 3:07 PM, Stéphane Rey wrote: Hi guys. After several experiments I could discover that the "bad" ADEV from the two GPSDO DUT are due to GPS lock losses. This is probably because the antenna is outside the windows but half the sky is hidden. We can see the on the frequency plot the sharp change of 0.5Hz and the locking. Good point. I'm now trying to evaluate various architectures of 2-channels squarers and a DMDT. For that I'm designing a PCB with 4 squarers : simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the transistor based differential amplifier from Winzel and the one from Charles. I will add two balanced mixers (minicircuits), IF filters and amplifiers. Does anyone has an idea of what I could add for this evaluation ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : mardi 20 janvier 2015 23:15 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi, Following the tests results in the previous email, today I've performed additional measurements showing that the repeatability of the GPSDO DUT is not great but is coming from the design. I've tested several over sources and repeatability is correct. I can already make some measurement. Good ! Now I'd like to improve. First I'm going to implement a squarer and then I will work on the DMTD... I'm thinking to make a setup on the table, and possibly make a small PCB then. Any comment for the tests results of yesterday here under ? Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Stéphane Rey Envoyé : lundi 19 janvier 2015 22:32 À : 'Discussion of precise time and frequency measurement' Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi Here are the results of today's experiments. plots and TIM files attached to this email. Setup #1 : dark blue I've done again the floor measurement with same conditions : HP58503 for 10 MHz Standard, 1PPS for the EXT gating an
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Stephane wrote: I'm now trying to evaluate various architectures of 2-channels squarers and a DMDT. For that I'm designing a PCB with 4 squarers : simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the transistor based differential amplifier from Winzel and the one from Charles. Note that squaring a 10MHz sine wave and squaring a 10 or 100Hz mixer output are two very different tasks. If you start at baseband, a Collins-style multi-stage limiting amp is a great benefit. That is generally not necessary if you start at 10MHz (or if you do use a Collins-style limiter it needs far fewer stages). All of the squarers you mention work well at 10MHz, but not as well at baseband. The LT1719 is easier to apply and faster than the LT1016. You may want to use that instead of the 1016. The LT1719 and LT1715 datasheets show the simplest possible implementation (see below). The MPSH81 devices in my version are available in surface-mount (MMBTH81) if that is more convenient. Other fast transistors will also work (BFT92, BFT93, BFG31). Best regards, Charles ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hi everyone. Many thanks for your very useful comments. I had already seen most of the documents you were pointing but not on the collins and Bruce discussion around the multistage filter. However I've already seen this approach in the document from Allan (http://tf.nist.gov/timefreq/general/pdf/84.pdf) At first I had in mind to square the 10 MHz but this is the aim of the evaluation board to evaluate various architectures. So I will implement several squarers including the Collins Approach both at 10 MHz and 100 Hz and all the blocks will have input and output connectors so that I will be able to test several layouts. I will show you the final design. Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Charles Steinmetz Envoyé : dimanche 25 janvier 2015 08:08 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Stephane wrote: >I'm now trying to evaluate various architectures of 2-channels squarers >and a DMDT. For that I'm designing a PCB with 4 squarers : >simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the transistor >based differential amplifier from Winzel and the one from Charles. Note that squaring a 10MHz sine wave and squaring a 10 or 100Hz mixer output are two very different tasks. If you start at baseband, a Collins-style multi-stage limiting amp is a great benefit. That is generally not necessary if you start at 10MHz (or if you do use a Collins-style limiter it needs far fewer stages). All of the squarers you mention work well at 10MHz, but not as well at baseband. The LT1719 is easier to apply and faster than the LT1016. You may want to use that instead of the 1016. The LT1719 and LT1715 datasheets show the simplest possible implementation (see below). The MPSH81 devices in my version are available in surface-mount (MMBTH81) if that is more convenient. Other fast transistors will also work (BFT92, BFT93, BFG31). Best regards, Charles --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.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] question Alan deviation measured with Timelab and counters
Hi The approach in the original NIST paper below was sort of a “best guess” about how to do the limiting and filtering. When the paper was presented, a number of us questioned how that part of the circuit was arrived at. The conversation more or less ended up with “that’s something we can investigate further”. The Collins paper (and Bruce’s work based on it) is a much better way to look at the 10 Hz squaring process. At 10 MHz, that stuff is not needed. Bob > On Jan 25, 2015, at 10:44 AM, Stéphane Rey wrote: > > Hi everyone. > > Many thanks for your very useful comments. > I had already seen most of the documents you were pointing but not on the > collins and Bruce discussion around the multistage filter. However I've > already seen this approach in the document from Allan > (http://tf.nist.gov/timefreq/general/pdf/84.pdf) > > At first I had in mind to square the 10 MHz but this is the aim of the > evaluation board to evaluate various architectures. So I will implement > several squarers including the Collins Approach both at 10 MHz and 100 Hz > and all the blocks will have input and output connectors so that I will be > able to test several layouts. > > I will show you the final design. > > Cheers > Stephane > > > -Message d'origine- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Charles > Steinmetz > Envoyé : dimanche 25 janvier 2015 08:08 > À : Discussion of precise time and frequency measurement > Objet : Re: [time-nuts] question Alan deviation measured with Timelab and > counters > > Stephane wrote: > >> I'm now trying to evaluate various architectures of 2-channels squarers >> and a DMDT. For that I'm designing a PCB with 4 squarers : >> simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the transistor >> based differential amplifier from Winzel and the one from Charles. > > Note that squaring a 10MHz sine wave and squaring a 10 or 100Hz mixer output > are two very different tasks. If you start at baseband, a Collins-style > multi-stage limiting amp is a great benefit. That is generally not > necessary if you start at 10MHz (or if you do use a Collins-style limiter it > needs far fewer stages). All of the squarers you mention work well at > 10MHz, but not as well at baseband. > > The LT1719 is easier to apply and faster than the LT1016. You may want to > use that instead of the 1016. The LT1719 and LT1715 datasheets show the > simplest possible implementation (see below). > > The MPSH81 devices in my version are available in surface-mount > (MMBTH81) if that is more convenient. Other fast transistors will also work > (BFT92, BFT93, BFG31). > > Best regards, > > Charles > > > > --- > L'absence de virus dans ce courrier électronique a été vérifiée par le > logiciel antivirus Avast. > http://www.avast.com > > ___ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] question Alan deviation measured with Timelab and counters
I do understand. Has anyone already compared the performances of squaring the 10 MHz vs squaring the IF ? Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Bob Camp Envoyé : dimanche 25 janvier 2015 19:01 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi The approach in the original NIST paper below was sort of a “best guess” about how to do the limiting and filtering. When the paper was presented, a number of us questioned how that part of the circuit was arrived at. The conversation more or less ended up with “that’s something we can investigate further”. The Collins paper (and Bruce’s work based on it) is a much better way to look at the 10 Hz squaring process. At 10 MHz, that stuff is not needed. Bob > On Jan 25, 2015, at 10:44 AM, Stéphane Rey wrote: > > Hi everyone. > > Many thanks for your very useful comments. > I had already seen most of the documents you were pointing but not on > the collins and Bruce discussion around the multistage filter. However > I've already seen this approach in the document from Allan > (http://tf.nist.gov/timefreq/general/pdf/84.pdf) > > At first I had in mind to square the 10 MHz but this is the aim of the > evaluation board to evaluate various architectures. So I will > implement several squarers including the Collins Approach both at 10 > MHz and 100 Hz and all the blocks will have input and output > connectors so that I will be able to test several layouts. > > I will show you the final design. > > Cheers > Stephane > > > -Message d'origine- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de > Charles Steinmetz Envoyé : dimanche 25 janvier 2015 08:08 À : > Discussion of precise time and frequency measurement Objet : Re: > [time-nuts] question Alan deviation measured with Timelab and counters > > Stephane wrote: > >> I'm now trying to evaluate various architectures of 2-channels >> squarers and a DMDT. For that I'm designing a PCB with 4 squarers : >> simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the >> transistor based differential amplifier from Winzel and the one from Charles. > > Note that squaring a 10MHz sine wave and squaring a 10 or 100Hz mixer > output are two very different tasks. If you start at baseband, a > Collins-style multi-stage limiting amp is a great benefit. That is > generally not necessary if you start at 10MHz (or if you do use a > Collins-style limiter it needs far fewer stages). All of the squarers > you mention work well at 10MHz, but not as well at baseband. > > The LT1719 is easier to apply and faster than the LT1016. You may > want to use that instead of the 1016. The LT1719 and LT1715 > datasheets show the simplest possible implementation (see below). > > The MPSH81 devices in my version are available in surface-mount > (MMBTH81) if that is more convenient. Other fast transistors will > also work (BFT92, BFT93, BFG31). > > Best regards, > > Charles > > > > --- > L'absence de virus dans ce courrier électronique a été vérifiée par le > logiciel antivirus Avast. > http://www.avast.com > > ___ > time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.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] question Alan deviation measured with Timelab and counters
The performance of the 2 systems should be comparable provided the similar equivalent noise bandwidths are used.Every 10Mhz edge needs to be timestamped with ps resolution and the resulting phase samples low pass filtered and decimated to achieve this.The 10MSPS picosecond or better resolution time stamping with femtosecond integral linearity will be a bit of a challenge to achieve. Bruce On Tuesday, 27 January 2015 3:26 PM, Stéphane Rey wrote: I do understand. Has anyone already compared the performances of squaring the 10 MHz vs squaring the IF ? Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Bob Camp Envoyé : dimanche 25 janvier 2015 19:01 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi The approach in the original NIST paper below was sort of a “best guess” about how to do the limiting and filtering. When the paper was presented, a number of us questioned how that part of the circuit was arrived at. The conversation more or less ended up with “that’s something we can investigate further”. The Collins paper (and Bruce’s work based on it) is a much better way to look at the 10 Hz squaring process. At 10 MHz, that stuff is not needed. Bob > On Jan 25, 2015, at 10:44 AM, Stéphane Rey wrote: > > Hi everyone. > > Many thanks for your very useful comments. > I had already seen most of the documents you were pointing but not on > the collins and Bruce discussion around the multistage filter. However > I've already seen this approach in the document from Allan > (http://tf.nist.gov/timefreq/general/pdf/84.pdf) > > At first I had in mind to square the 10 MHz but this is the aim of the > evaluation board to evaluate various architectures. So I will > implement several squarers including the Collins Approach both at 10 > MHz and 100 Hz and all the blocks will have input and output > connectors so that I will be able to test several layouts. > > I will show you the final design. > > Cheers > Stephane > > > -Message d'origine- > De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de > Charles Steinmetz Envoyé : dimanche 25 janvier 2015 08:08 À : > Discussion of precise time and frequency measurement Objet : Re: > [time-nuts] question Alan deviation measured with Timelab and counters > > Stephane wrote: > >> I'm now trying to evaluate various architectures of 2-channels >> squarers and a DMDT. For that I'm designing a PCB with 4 squarers : >> simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the >> transistor based differential amplifier from Winzel and the one from Charles. > > Note that squaring a 10MHz sine wave and squaring a 10 or 100Hz mixer > output are two very different tasks. If you start at baseband, a > Collins-style multi-stage limiting amp is a great benefit. That is > generally not necessary if you start at 10MHz (or if you do use a > Collins-style limiter it needs far fewer stages). All of the squarers > you mention work well at 10MHz, but not as well at baseband. > > The LT1719 is easier to apply and faster than the LT1016. You may > want to use that instead of the 1016. The LT1719 and LT1715 > datasheets show the simplest possible implementation (see below). > > The MPSH81 devices in my version are available in surface-mount > (MMBTH81) if that is more convenient. Other fast transistors will > also work (BFT92, BFT93, BFG31). > > Best regards, > > Charles > > > > --- > L'absence de virus dans ce courrier électronique a été vérifiée par le > logiciel antivirus Avast. > http://www.avast.com > > ___ > time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.com ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Thanks a lot for your comment Bruce, I need to feel a bit deeper the ins and outs of the methods so I guess I will anyway implement both methods on an evaluation PCB and characterize each method. This will bring to me some actual data to compare. I will share the results of course. The plan is to have an eval PCB with 4 independant 10 MHz squarers, isolation amplifiers, mixers, low pass filters and multistage limiting amplifier. Each block will have input/output connectors so that I can combine any architecture with these blocks. The PCB will receive a low noise PSU as well. Before I start the design if one thinks about something to add in the evaluation, this is very welcome. Stephane On Tue, 27 Jan 2015 03:24:44 + (UTC), Bruce Griffiths wrote: The performance of the 2 systems should be comparable provided the similar equivalent noise bandwidths are used.Every 10Mhz edge needs to be timestamped with ps resolution and the resulting phase samples low pass filtered and decimated to achieve this.The 10MSPS picosecond or better resolution time stamping with femtosecond integral linearity will be a bit of a challenge to achieve. Bruce On Tuesday, 27 January 2015 3:26 PM, Stéphane Rey wrote: I do understand. Has anyone already compared the performances of squaring the 10 MHz vs squaring the IF ? Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Bob Camp Envoyé : dimanche 25 janvier 2015 19:01 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Hi The approach in the original NIST paper below was sort of a “best guess” about how to do the limiting and filtering. When the paper was presented, a number of us questioned how that part of the circuit was arrived at. The conversation more or less ended up with “that’s something we can investigate further”. The Collins paper (and Bruce’s work based on it) is a much better way to look at the 10 Hz squaring process. At 10 MHz, that stuff is not needed. Bob On Jan 25, 2015, at 10:44 AM, Stéphane Rey wrote: Hi everyone. Many thanks for your very useful comments. I had already seen most of the documents you were pointing but not on the collins and Bruce discussion around the multistage filter. However I've already seen this approach in the document from Allan (http://tf.nist.gov/timefreq/general/pdf/84.pdf) At first I had in mind to square the 10 MHz but this is the aim of the evaluation board to evaluate various architectures. So I will implement several squarers including the Collins Approach both at 10 MHz and 100 Hz and all the blocks will have input and output connectors so that I will be able to test several layouts. I will show you the final design. Cheers Stephane -Message d'origine- De : time-nuts [mailto:time-nuts-boun...@febo.com] De la part de Charles Steinmetz Envoyé : dimanche 25 janvier 2015 08:08 À : Discussion of precise time and frequency measurement Objet : Re: [time-nuts] question Alan deviation measured with Timelab and counters Stephane wrote: I'm now trying to evaluate various architectures of 2-channels squarers and a DMDT. For that I'm designing a PCB with 4 squarers : simple 74ac04 gate biased at VCC/2, a LT1016 comparator, the transistor based differential amplifier from Winzel and the one from Charles. Note that squaring a 10MHz sine wave and squaring a 10 or 100Hz mixer output are two very different tasks. If you start at baseband, a Collins-style multi-stage limiting amp is a great benefit. That is generally not necessary if you start at 10MHz (or if you do use a Collins-style limiter it needs far fewer stages). All of the squarers you mention work well at 10MHz, but not as well at baseband. The LT1719 is easier to apply and faster than the LT1016. You may want to use that instead of the 1016. The LT1719 and LT1715 datasheets show the simplest possible implementation (see below). The MPSH81 devices in my version are available in surface-mount (MMBTH81) if that is more convenient. Other fast transistors will also work (BFT92, BFT93, BFG31). Best regards, Charles --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.com ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. --- L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast. http://www.avast.com ___ time-nuts ma
Re: [time-nuts] question Alan deviation measured with Timelab and counters
Moin Magnus, On Wed, 21 Jan 2015 07:07:54 +0100 Magnus Danielson wrote: > For oscillators, they should have been turned on long enough such that > any drift is negligible. Alternatively you process out the quadratic > trend out of it. The later should be accompanied by some quality measure > of how much remaining systematics there is (see Jim Barnes PTTI paper on > Drift Estimators). Do you mean "The measurement of linear frequency drift in oscillators", http://tf.nist.gov/timefreq/general/tn1337/Tn264.pdf ? Attila Kinali -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hej Attila, On 02/06/2015 01:22 PM, Attila Kinali wrote: Moin Magnus, On Wed, 21 Jan 2015 07:07:54 +0100 Magnus Danielson wrote: For oscillators, they should have been turned on long enough such that any drift is negligible. Alternatively you process out the quadratic trend out of it. The later should be accompanied by some quality measure of how much remaining systematics there is (see Jim Barnes PTTI paper on Drift Estimators). Do you mean "The measurement of linear frequency drift in oscillators", http://tf.nist.gov/timefreq/general/tn1337/Tn264.pdf ? Yes. I use the PTTI link http://tycho.usno.navy.mil/ptti/1983papers/Vol%2015_29.pdf I was just lazy to dig the link up when I wrote that. 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] question Alan deviation measured with Timelab and counters
On Tue, 20 Jan 2015 08:39:13 +0100 Attila Kinali wrote: > 2-3 years ago, i got a presentation of an italian amateur radio on > how to measure phase noise of osciallators using this technique in quite > detail. Only draw back is that the presentation is in italian. If someone > wants this presentation, please contact me off list. Azelio Boriani was so kind to translate the presentation. You can find the original pdfs and the Azelio's translations at http://attila.kinali.ch/phase_noise_measurement/ Attila Kinali -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ time-nuts mailing 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] question Alan deviation measured with Timelab and counters
Hi Wow, that guy did a *lot* of work on all that. Congratulations to the him and those kind enough to translate it. Per an off list request there’s at least one additional circuit to add to his list: Wire the RF as shown on page 17 of the MAdS slides but: 1) Use an RPD-1 (or modern equivalent) as the phase detector. 2) the 50 ohm on the filter input goes to 500 ohms. 3) Run an OP-37 (or modern equivalent) on +/- 18V 4) + input of the OP-37 goes straight to the lowpass filter output (Bruce has correctly recommend protection here if you don’t trust your supplies) 5) The - input of the OP-37 gets 100 ohms to ground and (roughly) 1K ohm to the output You now have a low(er) noise amp with fewer parts (other than the -18V supply). It also has no low frequency cutoff. That makes it useful for a few more things. There may / will be some peaking in the lowpass filter on the mixer. Exactly how much depends on the parts you use. You can either measure the peaking or model it. I would put blocking caps in between the mixer RF inputs and the J1 and J2. Mixers seem to last longer that way. John has suggested running blocking on both sides of the BNC’s to nuke a ground loop. That may also be a good idea. You might simply try running the mixer using it’s input transformers for isolation. That may or may not work depending on what’s really inside. === The other comment is that the “lock box” portion of the circuit for pages 17 and 18 is missing. The lock box drives the EFC on one of the OCXO’s you are testing. Without it you can’t really do a phase noise test. Yes, it’s not an exciting circuit, but you do need it to make the thing work. Given that the guy already had a massive number of topics to cover, it’s understandable that it’s missing. What you need: 1) A DC pickoff point from the preamp. In the circuit above, the output of the OP-37 is fine. 2) Some means of varying the gain. A pot or switched resistors both work. 3) An ability to crank in an offset. Again a pot works pretty well. 4) The ability to invert the signal if needed. The amp, pots, switches, drives the EFC on one of your test devices. You fiddle the offset to get them locked with zero DC on the preamp output. You fiddle the gain to get the low end of your measurement down far enough to be useful. I’ve seen people do this all with a couple of pots and a 9V battery. I typically do it with a pair of op-amps. You can add caps to the circuit. You can model the cutoff of the circuit. You can noise load and measure the impact of the circuit. All of these enhancements help at the low frequency end. None are needed if you are just starting out. Once you have this preamp (the OP-37 stuff) built up, it also works as the input for a single mixer ADEV setup. You need to put a limiter on the output, but there is no more RF work to do. With a little planning, you can have a phase noise and ADEV front end all in one box / all on one pc board. None of it is really critical. I’ve made them up a number of times on a chunk of copper clad. It will probably take you more time to build up a low noise / clean / reliable / ground isolated power supply than to make up the rest of the circuit. A ground loop through any of this will drive you a bit nuts. I would make up the power supply from scratch. Current required is nearly nothing (< 100 ma). You want an isolated supply, and if if faults you have big problems. Good old 7818 and 7918 regulators are plenty quiet enough. Op amps are good at rejecting audio frequency noise. Resist the temptation to hook anything else to the supply that feeds this box. Depending on the op-amps you pick, your supply may be +/-12 up to +/- 20V. The idea is to run as much gain as you can in the preamp without getting into clipping. The idea is *not* to blow out them out with a higher than rated supply. === Yes there’s a bit of calibration involved in getting this running. You can play with other mixers, they won’t be 500 ohm output. There is nothing special about an op amp that’s old enough to vote. This isn’t quite a single weekend project. It is if you have all the parts on hand and have done it a couple times. It should not be more than a couple weeks of fiddling to get one running. Cost wise, you should be able to do it for < $100 using perf board and copper clad. If you go the full up ADEV route that probably tacks on another few bucks and another weekend or two. 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.