Re: [time-nuts] 10MHz Rubidium reference source for frequency counter
Hi Tom, Thank you for the comments. The results are better than my expectation. Please find the new result with two Morion's OCXO: the first is an external reference for CNT-91 and second - DUT. https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/morion%20ocxo%20vs%20morion%20reference%20adev%20freq%20mode%20smartfreq%20sample_1s%2020141104.jpg Blue line - Frequency mode / Sample interval 1s / Smart Freq OFF / CNT-91 reference source - internal OCXO Crimson line - Frequency mode / Sample interval 1s / Smart Freq ON / CNT-91 reference source - internal OCXO Green line - Frequency mode / Sample interval 1s / Smart Freq ON / CNT-91 reference source - external Morion OCXO Regarding CNT-91 Max Measurement Rate (p.8-10 of CNT-91's User manual Rev.18): 1) via GPIB - 2000 readings/s (block) / 350 readings/s (individual). Hope it's OK for my Prologic USB-GPIB interface ... 2) To internal memory - 250 k readings/s One other thing to try is the ADEV function on the front panel. The CNT-91 is one of the few counters that calculates Allan deviation, not just mean and standard deviation. With this, you do not get a fancy log-log ADEV(tau) plot but the number reported will be valid for the gate time (tau) that you have selected. The advantage of this is that you don't need any GPIB adapter or computer connection or CNT-91-compatible PC software. So maybe try tau 1, 5, 10, 20, 50 and write down what it reports. Well, I use this feature (STAT/ADEV) also but only for the first time look. IMHO, if someone has GPIB interface TimeLab solution is more powerful and provides much more opportunities ... It also makes possible to document the results for further use. I didn't find how to change tau value for this mode and it seems by default tau is 1 sec. As I understand it, his project is to use the high frequency output of a ublox NEO-7M to discipline a MV89 with a James Miller-style analog PLL. This is a beginning of the project to get a compact and non-expensive GPSDO solution on base of uBlox GPS. The links are: http://www.ra3apw.ru/proekty/ublox-neo-7m/ http://www.ra3apw.ru/ublox-neo-7m-ocxo-gpsdo/ Karen Date: Mon, 3 Nov 2014 06:43:06 -0800 From: Tom Van Baak t...@leapsecond.com Hi Karen, Ok, you are making very good progress. Thanks for the update. Your new ADEV plot, Morion OCXO ADEV Freq mode SmartFreq Sample_1s InterpCalib_On.jpg is now looking more like I would expect; it is showing the expected trade-off in measurement speed and measurement resolution. I see you are getting below 5e-12, past about tau 10 seconds. There are some periodics in the measurement which might go away with different trigger levels or better isolation or much larger frequency delta between the external ref OCXO and external input OCXO. Attached is CNT91-error.gif -- the uncertainty calculation as found in chapter 8 of the CNT-91 manual. It is worth reading the details of this chapter to gain further understanding of the measurement process. The single-shot spec for this counter is 50 ps. But from the equation you can see why they are able to claim 1 ps rms resolution for sufficient averaging (up to 1000 measurements per second). This works because the CNT-91 is an extremely quick continuous time-stamping counter with massive internal memory (something like 750,000 readings in internal RAM). And it has true back-to-back (zero deadtime) period / frequency measurements, which are available over GPIB. The question is if these GPIB capabilities are compatible with TimeLab, which may just treat this as dumb frequency counter instead of a high speed timestamping counter. I will continue to test my own CNT-91 here to see if I can improve on the noise floor. I will loan John my CNT-91 if it looks like a change to TimeLab is necessary; to make best use of raw high-speed timestamp-based measurements. One other thing to try is the ADEV function on the front panel. The CNT-91 is one of the few counters that calculates Allan deviation, not just mean and standard deviation. With this, you do not get a fancy log-log ADEV(tau) plot but the number reported will be valid for the gate time (tau) that you have selected. The advantage of this is that you don't need any GPIB adapter or computer connection or CNT-91-compatible PC software. So maybe try tau 1, 5, 10, 20, 50 and write down what it reports. The other question is the validity of the ADEV calculation when the counter's smart features are turned on. The good news is that this should not impact your goal to compare multiple oscillators against each other; since that is essentially a pair-wise, relative comparison. But it may affect the absolute accuracy of the calculations themselves. Its a subtle point, which you are welcome to ignore for now. Yes, it is possible that some of the OCXO that you are testing are near or even better than the short-term measurement ability of the CNT-91. In this case either
Re: [time-nuts] 10MHz Rubidium reference source for frequency counter
Charles, I have a machine-translated PDF (~2.5MB), but nowhere to put it up. If you have web space for it, please feel free to do so (assuming that Karen does not object). I am OK with it if it might be interesting to someone... Karen Date: Mon, 03 Nov 2014 15:43:47 -0500 From: Charles Steinmetz csteinm...@yandex.com Tom wrote: As I understand it, his project is to use the high frequency output of a ublox NEO-7M to discipline a MV89 with a James Miller-style analog PLL. * * * (perhaps someone can post an English translation for us) Tom, I have a machine-translated PDF (~2.5MB), but nowhere to put it up. If you have web space for it, please feel free to do so (assuming that Karen does not object). I have reservations about the potential of any Miller-style GPSDO, because using an analog PLL essentially guarantees that the crossover to GPS will be at a tau orders of magnitude lower than it should be -- the Miller unit crosses over about 3 OOM too low (see attached plot, from http://www.leapsecond.com/pages/gpsdo/). The beloved Thunderbolt also crosses over about 2.5 OOM too early, if you use the factory loop settings (id.). But it has an ADPLL with adjustable parameters, so a time nut can tune the loop for best performance with the particular OCXO in the unit. (See http://www.ke5fx.com/tbolt.htm.) Best regards, Charles -- next part -- A non-text attachment was scrubbed... Name: Miller_GPSDO_ADEV.jpg Type: image/jpeg Size: 29695 bytes Desc: not available URL: http://www.febo.com/pipermail/time- nuts/attachments/20141103/a232b407/attachment.jpg -- Subject: Digest Footer ___ time-nuts mailing list time-nuts@febo.com https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts -- End of time-nuts Digest, Vol 124, Issue 35 ** ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hi, yes, it is my fault - my set up (connections) for TI mode was not correct. According your advices I went back to the frequency mode and set gate / sample time to 1 second for CNT-91 and TimeLab. Then I switched on Smart Frequency and Interpolator Calibration option on CNT-91 1. Setting - Misc - Smart Meas - Smart Frequency - ON Smart Frequency (valid only if the selected measurement function is Frequency or Period Average). By means of continuous timestamping and regression analysis, the resolution is increased for measuring times between 0.2 s and 100 s. 2. Setting - Misc - Smart Meas - Interp Calib ON Interpolatator Calibration - By switching off the interpolator calibration, you can increase the measurement speed at the expense of accuracy. Please see the result with and without these CNT-91's options (Smart Frequency and Interpolator Calibration) https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/morion%20ocxo%20adev%20freq%20mode%20smartfreq%20sample_1s%20interpcalib_on.jpg . This result is for the internal CNT-91 reference source. I will check with external reference second Morion OCXO shortly. Thanks, Karen ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Charles, John and Bob, Many thanks for your advices. 1) Charles Steinmetz wrote: I would suggest trying a few of the oscillators again (no need to try them all), using the Isotemp OCXO as the time base for the Pendulum by connecting it to the Pendulum's REF input and selecting External Reference from the front panel menus. This will tell you if there is a problem with the internal time base in the counter. Other external reference signal (from RS FSQ) for Pendulum CNT-91 tested. ADEV result is the same. If so CNT-91's internal OCXO is good enough. 2) Charles Steinmetz wrote: Are the input signals triggering the counter very stably near the middle of their peak-to-peak voltages? Checked that the input signals triggering the counter very stably near the middle of their peak-to-peak voltages. CNT-91 has an automatic trigger level set up and an indication of triggering. 3) Charles Steinmetz wrote: Does your sample interval give the counter time to process each new sample? As I saw TimeLab and CNT-91 set up a sample interval automatically when I press a Monitor button in TimeLab. I got a sample interval 0.04sec. 4) Charles Steinmetz wrote: Are you sure your TimeLab setup is correct? Not sure 100%. May be John can check - https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/ocxo%20adev%20ti%20mode%20config%2020141102.jpg 5) John Miles wrote: Also consider taking TI readings rather than frequency readings. Mode has been switch from Frequency to TI - result is much better now. It was the main improvement in my measurement. Thanks a lot! Picture of ADEV OCXO result - https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/OCXO%20ADEV%20TI%20mode%2020141102.jpg 6) Bob Camp wrote: One of the things that will drive them a bit nuts are sub-harmonics. If your OCXO’s are 5 MHz doubled to 10 MHz, they will have sub-harmonics (5 MHz, 15 MHz, 25 MHz etc). These “extra” signals are not what the counter is expecting to see on a 10 MHz signal. Harmonic level from OCXO checked - maximum harmonic levels (MV89A) are below -41 dBс 7) John Miles wrote: I have no experience with the Pendulum counters but I wouldn't expect them to do much better than 1E-10 at t=1s regardless of how you set up the measurement, just based on their single-shot resolution spec in the 50-100 ps region. I got this result 1E-10 at tau=1s with my tested OCXOs and my measure equipment. I am not sure that I need better measure stability for my current applications (WSPR, VHF/UHF synthesizers and LO). Thanks a lot for your advices. Karen, ra3apw Date: Fri, 31 Oct 2014 19:39:35 -0400 From: Charles Steinmetz csteinm...@yandex.com Karen wrote: All OCXOs have been tested in free-running mode after 40-50 min of switch-on/warming. All OCXOs have been installed in isolated close box to exclude external airflow. Cases of all OCXOs had not any contact with external metallic parts of the box. Then it appears that something is wrong with your method of measuring the ADEV. I would suggest trying a few of the oscillators again (no need to try them all), using the Isotemp OCXO as the time base for the Pendulum by connecting it to the Pendulum's REF input and selecting External Reference from the front panel menus. This will tell you if there is a problem with the internal time base in the counter. [NOTE that the REF input needs a sinewave between 0.1 Vrms and 5 Vrms. If the Isotemp has a square-wave output, choose another OCXO that fulfills the Pendulum specification.] If those tests produce ADEV measurements of 1e-11 or better at 1 second, the internal time base is too noisy for what you are trying to do. I expect that the internal time base will prove to be OK, but you need to verify that before proceeding. However, I note that in the manual, the internal time base of the OPT 19/90 counter is rated for a Root Allan Variance of 1e-10 at tau = 1 second and 10 seconds (see p. 8-15 of the 2014 User's Manual, Rev.18). OPT 30/90 is rated at 1e-11, and OPT 40/90 is rated at 5e-12. So, you may be getting about all of the performance the OPT 19/90 can deliver, if its OCXO just meets specification. (HP5370Bs are also specified at 1e-10, but they often deliver ADEV = 1e-11 or 1e-12 at 1 second. I do not have enough experience with Pendulum counters to know what kind of stability is typical.) Assuming that the test above (with the external REF) gives similar results to what you posted today, there is something wrong with your measurement setup. Are the input signals triggering the counter very stably near the middle of their peak-to-peak voltages? Is the counter adjusted to give maximum resolution for the sampling you are doing? Does your sample interval give the counter time to process each new sample? Have you read Chapters 3, 4, and 5 of the User's Manual carefully, and set up the counter properly for your measurements? Are you sure your
Re: [time-nuts] 10MHz Rubidium reference source for frequency counter
On Sun, 2 Nov 2014 12:13:34 +0300, you wrote: Mode has been switch from Frequency to TI - result is much better now. It was the main improvement in my measurement. Thanks a lot! Picture of ADEV OCXO result - https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/OCXO%20ADEV%20TI%20mode%2020141102.jpg Hi Karen, You appear just to be measuring the ADEV of your counter in single shot TI mode here, not the oscillators - how exactly do you have the counter and oscillators connected up? If I was doing it I'd probably try start by setting the counter to measure freq in Smart Freq mode, and set it to ext ref, with one morion feeding the external reference and another on the input. If the morions are ok you should get towards 1E-11 at 1 sec (if I'm reading the specs correctly) If not, try other oscillators and see if anything changes. As John said, TI is more complex to do anyway, and with quartz that has not had days or even weeks to settle it's a total pain. Angus ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hi The fact that the plots are still going down (as 1/tau) past the 1.0x10^-13 point at 1,000 seconds is also a pretty good indication that it's not really ADEV of an OCXO. The 100 ps rated single shot accuracy of the counter would give you 1.0 x10^-10 at 1 second and a 1/tau slope. Bob On Nov 2, 2014, at 3:56 PM, Angus not.ag...@btinternet.com wrote: On Sun, 2 Nov 2014 12:13:34 +0300, you wrote: Mode has been switch from Frequency to TI - result is much better now. It was the main improvement in my measurement. Thanks a lot! Picture of ADEV OCXO result - https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/OCXO%20ADEV%20TI%20mode%2020141102.jpg Hi Karen, You appear just to be measuring the ADEV of your counter in single shot TI mode here, not the oscillators - how exactly do you have the counter and oscillators connected up? If I was doing it I'd probably try start by setting the counter to measure freq in Smart Freq mode, and set it to ext ref, with one morion feeding the external reference and another on the input. If the morions are ok you should get towards 1E-11 at 1 sec (if I'm reading the specs correctly) If not, try other oscillators and see if anything changes. As John said, TI is more complex to do anyway, and with quartz that has not had days or even weeks to settle it's a total pain. Angus ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen wrote: Mode has been switch from Frequency to TI - result is much better now. It was the main improvement in my measurement. Your plot shows all traces starting at ~1e-9 at 0.1 second and dropping almost ruler-straight at 10x per decade. Any real crystal oscillator will flatten out somewhere between 1 second and 100 seconds, typically first to ~square root of 10 per decade, then to 0 per decade, and generally will be on the way back up before 1000 seconds and before getting to 1e-13. (For examples, see http://leapsecond.com/pages/z3801a-osc/.) Also, the fact that there is so little difference between the test oscillators is too unusual to be a coincidence. It does not appear that you are actually measuring the ADEV of the oscillators you are testing -- you seem to be measuring your test setup. 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] 10MHz Rubidium reference source for frequency counter
Mode has been switch from Frequency to TI - result is much better now. It was the main improvement in my measurement. Thanks a lot! Picture of ADEV OCXO result - https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/OCXO% 20ADEV%20TI%20mode%2020141102.jpg If you're taking a residual baseline measurement by feeding the same signal to both the start and stop inputs, then that looks OK. But if you're measuring one OCXO against another, then it wouldn't be correct since it doesn't show an increase in instability at longer taus. (If you like, send the .tim file to j...@miles.io and I'll have a look at it.) -- 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] 10MHz Rubidium reference source for frequency counter
John wrote: It seems to me that it's just a case of expecting too much from a counter. Possibly, but a well-tuned 5370B can get to the low e-12's at 1-10 seconds. It has better one-shot resolution than the Pendulum, but not by a factor of 10. So, I'd have thought the Pendulum would at least get to the mid e-11's if the time base was up to it. To get the best performance from a counter, you need to set up the trigger thresholds well away from ground as Charles says. Also consider taking TI readings rather than frequency readings. Frequency readings are OK for initial setup and basic ADEV comparisons but they won't generally yield the best ADEV fidelity or the lowest possible measurement floor. Yes, using TI mode is essential for getting down to the counter's limits. Was Karen using frequency mode? I wasn't paying attention to that end of things. 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] 10MHz Rubidium reference source for frequency counter
csteinm...@yandex.com said: Yes, using TI mode is essential for getting down to the counter's limits. What's going on there? It's just a divide, right? Is the firmware not smart enough to do get enough precision? Do all counters have that problem? -- These are my opinions. I hate spam. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 10MHz Rubidium reference source for frequency counter
It seems to me that it's just a case of expecting too much from a counter. Possibly, but a well-tuned 5370B can get to the low e-12's at 1-10 seconds. You can get down there with TI averaging, but the data you get is not ideal since the averaging process smooths out the very instabilities you're trying to characterize. (See Tom's page at http://www.leapsecond.com/pages/adev-avg/ for a good intuitive explanation.) Yes, using TI mode is essential for getting down to the counter's limits. Was Karen using frequency mode? I wasn't paying attention to that end of things. I'm not sure, but that's where most people start out. Going to TI mode may improve things a bit at the cost of a more complex, error-prone measurement process, but it still isn't going to measure a Morion at t=1s. -- 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] 10MHz Rubidium reference source for frequency counter
On Fri, 31 Oct 2014 23:40:04 -0400, you wrote: John wrote: It seems to me that it's just a case of expecting too much from a counter. Possibly, but a well-tuned 5370B can get to the low e-12's at 1-10 seconds. It has better one-shot resolution than the Pendulum, but not by a factor of 10. So, I'd have thought the Pendulum would at least get to the mid e-11's if the time base was up to it. The 'Smart Freq' feature on the CNT-91 appears to do multiple measurements over the gate time to get an improved resolution, kind of like the 53131/2. In practice, when 11 digits or whatever are claimed, I usually find that there is just a little bit of information in the last digit, not that it's a solid reading to that number of digits. In the Measurement Uncertainties section of the manual, would the formula for frequency suggest that Smart Freq makes the result up from multiple measurements of 0.4 times the nominal gate time? To get the best performance from a counter, you need to set up the trigger thresholds well away from ground as Charles says. Also consider taking TI readings rather than frequency readings. Frequency readings are OK for initial setup and basic ADEV comparisons but they won't generally yield the best ADEV fidelity or the lowest possible measurement floor. Yes, using TI mode is essential for getting down to the counter's limits. Was Karen using frequency mode? I wasn't paying attention to that end of things. I don't know exactly how the CNT-91 does it, but the older CNT-81 can do a lot of measurements in Time Interval A to B mode. With a static delay, my 6681 gives a SD of about 0.22ps for a set of 100 1-second measurements (a 10MHz square wave was driving the inputs). One little thing I noticed there was that the reference freq needed to be at least 3ppm off the input frequency to get quite that low. The newer series seem to need a bit of an offset too: https://www.febo.com/pipermail/time-nuts/2014-September/086701.html Angus. 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. ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hal wrote: Yes, using TI mode is essential for getting down to the counter's limits. What's going on there? It's just a divide, right? Is the firmware not smart enough to do get enough precision? Do all counters have that problem? The raw TI data have all of the benefit of the interpolation done by the basic TI counter. To get frequency (or period), the counter divides the TI into (or by) the event count, which is an integer that is more granular than the interpolated TI result. This reduces precision (not necessarily by a huge amount, since it's typically a big integer) and has the potential to introduce periodic anomalies into the data. To my knowledge, all TI-based counters behave this way. 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] 10MHz Rubidium reference source for frequency counter
John, Thank you, done. I saved all .TIM files for each OCXO to have an opportunity to compare all variants off line. I will publish the results shortly. Karen Date: Wed, 29 Oct 2014 13:43:51 -0700 From: John Miles j...@miles.io Don't forget to hit 'f' and/or 'p' to check the frequency and phase traces when you're trying to understand what went wrong with an ADEV plot. Under the right (wrong) conditions, it only takes one or two outlying data points caused by power glitches or any number of other unrelated issues to degrade the whole plot. -- 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] 10MHz Rubidium reference source for frequency counter
Alex, Thank you for the advice. I have a colleague in Moscow who has a big experience with MV89A repair. 73 de Karen, ra3apw Date: Wed, 29 Oct 2014 14:57:17 -0700 From: Alexander Pummer alex...@ieee.org Don't give up Karen, the Morion oxcos are relative easy to repair, you need one coca-cola can, cut it is pieces until you get a very thin ALU foil, get some thicker carton wind around the box of the OXCO's box that way you have a heat isolation between the OXCO's box and the vise, which you need to hold it, but hold the OXCO that way, that the bottom part where the soldering is fare away from the vise, and the can is upside down, fire up a relative large [ min. 100W] solder iron, if it has temperature setting set to the maximum and if it is hot push it hard against the OXCO can, as soon as the solder melted push a piece Alu foil into the gap [ there is a gap between the bottom of the can and the rst of the can ] and move to the next part to melt the solder. You need to push in at least 5mm long the Alu foil into the gap, if you get it all around, you could pull out the content of the OXCO can. That is a double oven oscillator, depend what is the problem , you may have to open the internal can also, usually it is visible if something is broken, just power up and follow the current paths, with meter and scope probe, it is practical to temporarily disconnect the oven heating [ the large heater transistor ]. I fixed five of them, you need some patience and luck and you have to be very careful. 73 KJ6UHN Alex ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Thanks All for your comments and clarifications. The next OCXOs have been tested today: * Morion MV89A 10.000MHz #1 * Morion MV89A 10.000MHz #2 * Morion MV89A 10.000MHz #3 * Isotemp 131-191 10.000MHz #1 * Pilot C3Z 10.000MHz * Bliley NV26R891 100.000MHz * Giacint-M 5.000MHz Measurement set up: * Pendulum CNT-91 with TimeBase opt.19 and 10 * Prologix GPIB-USB * TimeLab V1.2 (x64) * Windows7 Pro (x64) * PC i5 /3.4GHz / 16GB ADEV results https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/ocxos_adev_test_20141031.jpg TimeLab configuration https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/timelab_adev_config.jpg All OCXO's TimeLab TIM files can be download https://dl.dropboxusercontent.com/u/21338179/Hamradio/OCXO_Adev/OCXO_Adev.zip A good MV89A can be below 1E-12. As you can see ADEV of all my tested OCXOs are not better than 4E(-10) for tau =1s. Question: can I be more or less happy with this result or something wrong in my measurement? Karen Date: Thu, 30 Oct 2014 23:32:48 +0100 From: Adrian rfn...@arcor.de Karen, as by the datasheet, the time interval resolution of your counter is 50 ps. This translates into an ADEV value of 5E-11 at tau = 1 sec, 5E-12 at tau = 10 sec etc, and represents the theoretical measurement limit of the counter. It can be measured by feeding the same signal into both inputs (start and stop) at the same time. A short delay (additional length of coaxial cable) at the stop input keeps the counter from producing phase wraps (jumping between a full input signal period and zero). A good MV89A can be below 1E-12. For measuring two good MV89A's against each other, a TI resolution of 50 ps is insufficient because up to 100 sec, to the most part of it, the actual ADEV of the oscillators can be below the measurement limit of the counter. Adrian ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen wrote: The next OCXOs have been tested today: Did you test the oscillators free-running (NOT in your GPSDO)? If so, those plots look much too high (by at least an order of magnitude, maybe even 2 or more OOM for some of the oscillators). Of course, oscillators take time to settle to their best stability -- as much as several weeks or even several months for some -- so if you just turned them on and measured them a few hours later you can expect higher than normal ADEV results. Even so, these results seem high. But if you measured them in your GPSDO circuit, you have not separated the potential measurement problems from potential GPSDO problems. If you suspect the time base in the Pendulum, you can use one of the OCXOs to feed its external reference input. 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] 10MHz Rubidium reference source for frequency counter
Thank you, Charles. All OCXOs have been tested in free-running mode after 40-50 min of switch-on/warming. All OCXOs have been installed in isolated close box to exclude external airflow. Cases of all OCXOs had not any contact with external metallic parts of the box. Karen Karen wrote: The next OCXOs have been tested today: Did you test the oscillators free-running (NOT in your GPSDO)? If so, those plots look much too high (by at least an order of magnitude, maybe even 2 or more OOM for some of the oscillators). Of course, oscillators take time to settle to their best stability -- as much as several weeks or even several months for some -- so if you just turned them on and measured them a few hours later you can expect higher than normal ADEV results. Even so, these results seem high. But if you measured them in your GPSDO circuit, you have not separated the potential measurement problems from potential GPSDO problems. If you suspect the time base in the Pendulum, you can use one of the OCXOs to feed its external reference input. 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] 10MHz Rubidium reference source for frequency counter
Karen wrote: All OCXOs have been tested in free-running mode after 40-50 min of switch-on/warming. All OCXOs have been installed in isolated close box to exclude external airflow. Cases of all OCXOs had not any contact with external metallic parts of the box. Then it appears that something is wrong with your method of measuring the ADEV. I would suggest trying a few of the oscillators again (no need to try them all), using the Isotemp OCXO as the time base for the Pendulum by connecting it to the Pendulum's REF input and selecting External Reference from the front panel menus. This will tell you if there is a problem with the internal time base in the counter. [NOTE that the REF input needs a sinewave between 0.1 Vrms and 5 Vrms. If the Isotemp has a square-wave output, choose another OCXO that fulfills the Pendulum specification.] If those tests produce ADEV measurements of 1e-11 or better at 1 second, the internal time base is too noisy for what you are trying to do. I expect that the internal time base will prove to be OK, but you need to verify that before proceeding. However, I note that in the manual, the internal time base of the OPT 19/90 counter is rated for a Root Allan Variance of 1e-10 at tau = 1 second and 10 seconds (see p. 8-15 of the 2014 User's Manual, Rev.18). OPT 30/90 is rated at 1e-11, and OPT 40/90 is rated at 5e-12. So, you may be getting about all of the performance the OPT 19/90 can deliver, if its OCXO just meets specification. (HP5370Bs are also specified at 1e-10, but they often deliver ADEV = 1e-11 or 1e-12 at 1 second. I do not have enough experience with Pendulum counters to know what kind of stability is typical.) Assuming that the test above (with the external REF) gives similar results to what you posted today, there is something wrong with your measurement setup. Are the input signals triggering the counter very stably near the middle of their peak-to-peak voltages? Is the counter adjusted to give maximum resolution for the sampling you are doing? Does your sample interval give the counter time to process each new sample? Have you read Chapters 3, 4, and 5 of the User's Manual carefully, and set up the counter properly for your measurements? Are you sure your TimeLab setup is correct? (I'll leave it to John to comment on that.) 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] 10MHz Rubidium reference source for frequency counter
Assuming that the test above (with the external REF) gives similar results to what you posted today, there is something wrong with your measurement setup. Are the input signals triggering the counter very stably near the middle of their peak-to-peak voltages? Is the counter adjusted to give maximum resolution for the sampling you are doing? Does your sample interval give the counter time to process each new sample? Have you read Chapters 3, 4, and 5 of the User's Manual carefully, and set up the counter properly for your measurements? Are you sure your TimeLab setup is correct? (I'll leave it to John to comment on that.) It seems to me that it's just a case of expecting too much from a counter. This is what a DMTD is for (or any of several other types of instruments ranging from easy homebrew projects to pricy commercial products.) To get the best performance from a counter, you need to set up the trigger thresholds well away from ground as Charles says. Also consider taking TI readings rather than frequency readings. Frequency readings are OK for initial setup and basic ADEV comparisons but they won't generally yield the best ADEV fidelity or the lowest possible measurement floor. I have no experience with the Pendulum counters but I wouldn't expect them to do much better than 1E-10 at t=1s regardless of how you set up the measurement, just based on their single-shot resolution spec in the 50-100 ps region. By itself, a counter is a good tool for monitoring long-term performance (meaning minutes to days), but it's just not the right tool for measuring the short-term stability of a quality HF OCXO. For your (Karen's) task, I would suggest taking a look at the tight PLL methodology -- see http://www.ke5fx.com/tpll.htm and other references linked from that page. I think that's probably the easiest way to measure devices in the 1E-12/s range. It's an underrated, underutilized technique. -- 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] 10MHz Rubidium reference source for frequency counter
Hi Unless the OCXO’s you have are broken, there is something wrong with your measurement. I have bought enough broken OCXO’s to know that you can get quite a few bad ones for every good one you find. The CNT counters are good instruments. They will do about as well as any frequency counter out there. They will not get to 1.0 x 10^-12 at one second. Depending on your signals and your setup, they may not get much past 1.0 x 10^-10 at one second. One of the things that will drive them a bit nuts are sub-harmonics. If your OCXO’s are 5 MHz doubled to 10 MHz, they will have sub-harmonics (5 MHz, 15 MHz, 25 MHz etc). These “extra” signals are not what the counter is expecting to see on a 10 MHz signal. Since your measurements are different for each OCXO, and go down by square root of tau, it’s got to be something unusual. The interaction of sub-harmonics with the CNT series frequency approximation algorithm may be it. Bob On Oct 31, 2014, at 3:57 PM, Karen Tadevosyan ra3...@mail.ru wrote: Thanks All for your comments and clarifications. The next OCXOs have been tested today: * Morion MV89A 10.000MHz #1 * Morion MV89A 10.000MHz #2 * Morion MV89A 10.000MHz #3 * Isotemp 131-191 10.000MHz #1 * Pilot C3Z 10.000MHz * Bliley NV26R891 100.000MHz * Giacint-M 5.000MHz Measurement set up: * Pendulum CNT-91 with TimeBase opt.19 and 10 * Prologix GPIB-USB * TimeLab V1.2 (x64) * Windows7 Pro (x64) * PC i5 /3.4GHz / 16GB ADEV results https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/ocxos_adev_test_20141031.jpg TimeLab configuration https://dl.dropboxusercontent.com/u/21338179/hamradio/OCXO_Adev/timelab_adev_config.jpg All OCXO's TimeLab TIM files can be download https://dl.dropboxusercontent.com/u/21338179/Hamradio/OCXO_Adev/OCXO_Adev.zip A good MV89A can be below 1E-12. As you can see ADEV of all my tested OCXOs are not better than 4E(-10) for tau =1s. Question: can I be more or less happy with this result or something wrong in my measurement? Karen Date: Thu, 30 Oct 2014 23:32:48 +0100 From: Adrian rfn...@arcor.de Karen, as by the datasheet, the time interval resolution of your counter is 50 ps. This translates into an ADEV value of 5E-11 at tau = 1 sec, 5E-12 at tau = 10 sec etc, and represents the theoretical measurement limit of the counter. It can be measured by feeding the same signal into both inputs (start and stop) at the same time. A short delay (additional length of coaxial cable) at the stop input keeps the counter from producing phase wraps (jumping between a full input signal period and zero). A good MV89A can be below 1E-12. For measuring two good MV89A's against each other, a TI resolution of 50 ps is insufficient because up to 100 sec, to the most part of it, the actual ADEV of the oscillators can be below the measurement limit of the counter. Adrian ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen, as by the datasheet, the time interval resolution of your counter is 50 ps. This translates into an ADEV value of 5E-11 at tau = 1 sec, 5E-12 at tau = 10 sec etc, and represents the theoretical measurement limit of the counter. It can be measured by feeding the same signal into both inputs (start and stop) at the same time. A short delay (additional length of coaxial cable) at the stop input keeps the counter from producing phase wraps (jumping between a full input signal period and zero). A good MV89A can be below 1E-12. For measuring two good MV89A's against each other, a TI resolution of 50 ps is insufficient because up to 100 sec, to the most part of it, the actual ADEV of the oscillators can be below the measurement limit of the counter. Adrian Karen Tadevosyan schrieb: Thanks again for your explanations and advices. I raised this question as a ADEV measurement result of my GPSDO/NEO-7M is about 1E-9 for tau = 1sec http://www.ra3apw.ru/wp-content/uploads/2014/10/NEO7M_OCXO_MV89A_10M_2h0_ADEV.jpg and it isn't meet my expectation. In my understanding (before I published my question here) the reason for this is not a good enough short range stability of my frequency counter's reference OCXO that's why I had an idea to use external rubidium source. Now I see that even my internal reference OCXO provides a good ADEV value 1E-10 for tau =1 sec (according Pendulum CNT-91 with opt.19 datasheet). If so what can be a reason of not a good ADEV result of GPSDO/NEO-7M? Karen, ra3apw ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hi The gotcha with translating ADEV directly to a measurement is the nature of ADEV. You are looking at a standard deviation measure, so the result will be some sort of “one sigma” kind of measure. Bob On Oct 30, 2014, at 6:32 PM, Adrian rfn...@arcor.de wrote: Karen, as by the datasheet, the time interval resolution of your counter is 50 ps. This translates into an ADEV value of 5E-11 at tau = 1 sec, 5E-12 at tau = 10 sec etc, and represents the theoretical measurement limit of the counter. It can be measured by feeding the same signal into both inputs (start and stop) at the same time. A short delay (additional length of coaxial cable) at the stop input keeps the counter from producing phase wraps (jumping between a full input signal period and zero). A good MV89A can be below 1E-12. For measuring two good MV89A's against each other, a TI resolution of 50 ps is insufficient because up to 100 sec, to the most part of it, the actual ADEV of the oscillators can be below the measurement limit of the counter. Adrian Karen Tadevosyan schrieb: Thanks again for your explanations and advices. I raised this question as a ADEV measurement result of my GPSDO/NEO-7M is about 1E-9 for tau = 1sec http://www.ra3apw.ru/wp-content/uploads/2014/10/NEO7M_OCXO_MV89A_10M_2h0_ADEV.jpg and it isn't meet my expectation. In my understanding (before I published my question here) the reason for this is not a good enough short range stability of my frequency counter's reference OCXO that's why I had an idea to use external rubidium source. Now I see that even my internal reference OCXO provides a good ADEV value 1E-10 for tau =1 sec (according Pendulum CNT-91 with opt.19 datasheet). If so what can be a reason of not a good ADEV result of GPSDO/NEO-7M? Karen, ra3apw ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen wrote: I raised this question as a ADEV measurement result of my GPSDO/NEO-7M is about 1E-9 for tau = 1sec and it isn't meet my expectation. In my understanding (before I published my question here) the reason for this is not a good enough short range stability of my frequency counter's reference OCXO that's why I had an idea to use external rubidium source. The Pendulum is a well-respected instrument for this sort of measurement, as you have now discovered by reading the datasheet. Much better than the ADEV you are getting. (In fact, it is probably better than the datasheet specification.) Now I see that even my internal reference OCXO provides a good ADEV value 1E-10 for tau =1 sec (according Pendulum CNT-91 with opt.19 datasheet). If so what can be a reason of not a good ADEV result of GPSDO/NEO-7M? (1) poor performance of the particular OCXO you have, or (2) poor GPSDO design or construction. To check (1), remove the OCXO from the GPSDO and measure its ADEV by itself. A good OCXO should be better that 1e-11 at 1 second, certainly better than 1e-10. I believe you said you are using a Morion MV-89. Many of the surplus MV-89s available in the US are broken and have poor ADEV or lots of spurs. The situation may be better where you are. If the ADEV of the oscillator by itself is only 1e-9, then you need [at least] a different oscillator. If the oscillator by itself is OK, then the GPSDO is not doing a good job of disciplining it. The purpose of a GPSDO is to let the oscillator run without interference at tau where the oscillator stability is better than the GPS timing signal, then to cross over at longer tau and let the GPS control the oscillator at averaging times where the GPS stability is better than the oscillator. Usually, this will be somewhere between 100 and 1000 seconds. So the control loop must be VERY slow, with a time constant in the hundreds or thousands of seconds -- it should be doing essentially nothing at tau = 1 second. The ADEV of the GPSDO at 1 second should be the same as the ADEV of the OCXO at 1 second, or very close to it. 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] 10MHz Rubidium reference source for frequency counter
Le 29 oct. 2014 à 08:32, Karen Tadevosyan a écrit : Thanks again for your explanations and advices. I raised this question as a ADEV measurement result of my GPSDO/NEO-7M is about 1E-9 for tau = 1sec http://www.ra3apw.ru/wp-content/uploads/2014/10/NEO7M_OCXO_MV89A_10M_2h0_ADEV.jpg and it isn't meet my expectation. In my understanding (before I published my question here) the reason for this is not a good enough short range stability of my frequency counter's reference OCXO that's why I had an idea to use external rubidium source. Now I see that even my internal reference OCXO provides a good ADEV value 1E-10 for tau =1 sec (according Pendulum CNT-91 with opt.19 datasheet). If so what can be a reason of not a good ADEV result of GPSDO/NEO-7M? Without GPS steering, the oscillator that you have should be able to give you low 10^12 stability at tau 1s once it has been running for a few days. Have you measured it to see if it is reasonably stable when free running?, at least with a consistent drift. If it is not then GPS won't help. If it is a good rock , then try using a pot vary the frequency. If it won't steer then there maybe something wrong in the box, or if the osc is old, maybe you have hit the min/max adjustment limits. Have you monitored your EFC corrections? You may have a too short time constant or too course adjustments causing overshoots. Karen, ra3apw ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Charles, Thanks a lot for the detailed explanation and recipes. Understood and will check this weekend. All my Morion 10MHz MV89A OCXOs are used and from eBay's. That's why the first version of broken surplus MV89A looks realistic. Best regards, Karen Date: Wed, 29 Oct 2014 10:02:44 -0400 From: Charles Steinmetz csteinm...@yandex.com Karen wrote: I raised this question as a ADEV measurement result of my GPSDO/NEO-7M is about 1E-9 for tau = 1sec and it isn't meet my expectation. In my understanding (before I published my question here) the reason for this is not a good enough short range stability of my frequency counter's reference OCXO that's why I had an idea to use external rubidium source. The Pendulum is a well-respected instrument for this sort of measurement, as you have now discovered by reading the datasheet. Much better than the ADEV you are getting. (In fact, it is probably better than the datasheet specification.) Now I see that even my internal reference OCXO provides a good ADEV value 1E-10 for tau =1 sec (according Pendulum CNT-91 with opt.19 datasheet). If so what can be a reason of not a good ADEV result of GPSDO/NEO-7M? (1) poor performance of the particular OCXO you have, or (2) poor GPSDO design or construction. To check (1), remove the OCXO from the GPSDO and measure its ADEV by itself. A good OCXO should be better that 1e-11 at 1 second, certainly better than 1e-10. I believe you said you are using a Morion MV-89. Many of the surplus MV-89s available in the US are broken and have poor ADEV or lots of spurs. The situation may be better where you are. If the ADEV of the oscillator by itself is only 1e-9, then you need [at least] a different oscillator. If the oscillator by itself is OK, then the GPSDO is not doing a good job of disciplining it. The purpose of a GPSDO is to let the oscillator run without interference at tau where the oscillator stability is better than the GPS timing signal, then to cross over at longer tau and let the GPS control the oscillator at averaging times where the GPS stability is better than the oscillator. Usually, this will be somewhere between 100 and 1000 seconds. So the control loop must be VERY slow, with a time constant in the hundreds or thousands of seconds -- it should be doing essentially nothing at tau = 1 second. The ADEV of the GPSDO at 1 second should be the same as the ADEV of the OCXO at 1 second, or very close to it. 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] 10MHz Rubidium reference source for frequency counter
Charles, Thanks a lot for the detailed explanation and recipes. Understood and will check this weekend. All my Morion 10MHz MV89A OCXOs are used and from eBay's. That's why the first version of broken surplus MV89A looks realistic. Best regards, Karen Don't forget to hit 'f' and/or 'p' to check the frequency and phase traces when you're trying to understand what went wrong with an ADEV plot. Under the right (wrong) conditions, it only takes one or two outlying data points caused by power glitches or any number of other unrelated issues to degrade the whole plot. -- 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] 10MHz Rubidium reference source for frequency counter
Don't give up Karen, the Morion oxcos are relative easy to repair, you need one coca-cola can, cut it is pieces until you get a very thin ALU foil, get some thicker carton wind around the box of the OXCO's box that way you have a heat isolation between the OXCO's box and the vise, which you need to hold it, but hold the OXCO that way, that the bottom part where the soldering is fare away from the vise, and the can is upside down, fire up a relative large [ min. 100W] solder iron, if it has temperature setting set to the maximum and if it is hot push it hard against the OXCO can, as soon as the solder melted push a piece Alu foil into the gap [ there is a gap between the bottom of the can and the rst of the can ] and move to the next part to melt the solder. You need to push in at least 5mm long the Alu foil into the gap, if you get it all around, you could pull out the content of the OXCO can. That is a double oven oscillator, depend what is the problem , you may have to open the internal can also, usually it is visible if something is broken, just power up and follow the current paths, with meter and scope probe, it is practical to temporarily disconnect the oven heating [ the large heater transistor ]. I fixed five of them, you need some patience and luck and you have to be very careful. 73 KJ6UHN Alex On 10/29/2014 11:09 AM, Karen Tadevosyan wrote: Charles, Thanks a lot for the detailed explanation and recipes. Understood and will check this weekend. All my Morion 10MHz MV89A OCXOs are used and from eBay's. That's why the first version of broken surplus MV89A looks realistic. Best regards, Karen ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hi Straight sawtooth corrected GPS starts out with (say) a 1x10^-9 ADEV at 1 second. That ADEV likely drops as 1/tau. If you go to 10 seconds it’s 1x10^-10. Your Rb starts out with a (say) 1x10^-11 ADEV at 1 second. That ADEV likely drops as 1/ square root (tau). If you go to 100 seconds it drops to 1x10^-12 Your OCXO starts out with a (say) 1x10^-12 ADEV at 1 second. That ADEV likely is the same as tau goes up to 10 or 100 seconds. Both the OCXO and Rb are sensitive to temperature. They may change by may parts in 10^12 per degree or per ten degrees. It depends a lot on how good a unit you have. In a room with a lot of drafts (temperature changes) this may impact ADEV even at 100 seconds. Nothing ever gets better and better forever. All of these standards level out at some point. GPS is unlikely to get past 1x10^-14. Most Rb’s struggle to get to 1x10^-13. An OCXO is unusual if it’s below 1x10^-12 at 1,000 seconds. No matter what sort of standard you have, it will perform better if you keep it powered up all the time. Bob On Oct 29, 2014, at 2:09 PM, Karen Tadevosyan ra3...@mail.ru wrote: Charles, Thanks a lot for the detailed explanation and recipes. Understood and will check this weekend. All my Morion 10MHz MV89A OCXOs are used and from eBay's. That's why the first version of broken surplus MV89A looks realistic. Best regards, Karen Date: Wed, 29 Oct 2014 10:02:44 -0400 From: Charles Steinmetz csteinm...@yandex.com Karen wrote: I raised this question as a ADEV measurement result of my GPSDO/NEO-7M is about 1E-9 for tau = 1sec and it isn't meet my expectation. In my understanding (before I published my question here) the reason for this is not a good enough short range stability of my frequency counter's reference OCXO that's why I had an idea to use external rubidium source. The Pendulum is a well-respected instrument for this sort of measurement, as you have now discovered by reading the datasheet. Much better than the ADEV you are getting. (In fact, it is probably better than the datasheet specification.) Now I see that even my internal reference OCXO provides a good ADEV value 1E-10 for tau =1 sec (according Pendulum CNT-91 with opt.19 datasheet). If so what can be a reason of not a good ADEV result of GPSDO/NEO-7M? (1) poor performance of the particular OCXO you have, or (2) poor GPSDO design or construction. To check (1), remove the OCXO from the GPSDO and measure its ADEV by itself. A good OCXO should be better that 1e-11 at 1 second, certainly better than 1e-10. I believe you said you are using a Morion MV-89. Many of the surplus MV-89s available in the US are broken and have poor ADEV or lots of spurs. The situation may be better where you are. If the ADEV of the oscillator by itself is only 1e-9, then you need [at least] a different oscillator. If the oscillator by itself is OK, then the GPSDO is not doing a good job of disciplining it. The purpose of a GPSDO is to let the oscillator run without interference at tau where the oscillator stability is better than the GPS timing signal, then to cross over at longer tau and let the GPS control the oscillator at averaging times where the GPS stability is better than the oscillator. Usually, this will be somewhere between 100 and 1000 seconds. So the control loop must be VERY slow, with a time constant in the hundreds or thousands of seconds -- it should be doing essentially nothing at tau = 1 second. The ADEV of the GPSDO at 1 second should be the same as the ADEV of the OCXO at 1 second, or very close to it. 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. ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Bert, Thank you for the interesting GPSDO/FE5680A proposal. I will contact you off list. Karen, ra3apw Date: Sun, 26 Oct 2014 17:22:45 -0400 From: ewkeh...@aol.com Karen To a large degree it depends on what you want to use it for. HP 5065 is considered top of the line and PRS 10 is a very nice Rb but lately I have seen lamp oscillator failures. I like FRK, well documented easy to modify and if done right super performance. As part of a GPSDO project using the FE 5680A that my Swiss friend Juerg and I have worked on for a year and is now in Beta test is also an alternative. We are done with FE 5680A so Juerg has his along with a GPSDO for sale. If you want a affordable Rb that has GPS control that is an alternative. Unit has less than 6 month running time so like new. Bert Kehren in Miami ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
John, Thank you for the links (I visited few times before) and your recommendations. Really I am looking for a source with a good short range stability performance between 1 and 10 seconds. I will prepare additional description of my task with more details shortly ... Special Thanks for your TimeLab software which I use for ADEV and PN measurements. Karen, ra3apw Date: Sun, 26 Oct 2014 14:28:24 -0700 From: John Miles j...@miles.io Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. You might take a look at the plots at http://www.ke5fx.com/rb.htm if you haven't already, and also at http://febo.com/pages/oscillators/rubes/ . These will give you an idea what to expect from some of the popular Rb models. The PRS10 mentioned by Said is a great choice for long-term measurements, but it's also the most expensive of the ones you are likely to find on the surplus market. Interestingly, if you want the best performance between 1 and 10 seconds the PRS10 is actually the worst of the ones I measured. But the CNT-91 probably will not be able to see the difference, since its single-shot resolution is a bit worse than the HP 5370B counter (see attached plot). Caveats: 1) No rubidium standard will give its best long-term performance until it has been allowed to run undisturbed for at least a few days. 2) These plots shouldn't be taken too seriously because of the (very) small number of examples tested under inconsistent environmental conditions. -- 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] 10MHz Rubidium reference source for frequency counter
Hi Charles: The SRS PRS-10 is based on their 10 MHz SC-10 oscillator. This is what I installed in the Gibbs rack box and made use of it's the power supplies and added down counters to get a 1 PPS output. http://www.thinksrs.com/products/SC10.htm There are options for phase noise and aging and I suspect these same options could be used when ordering the PRS-10. Maybe Tom has data on some of the SC-10 oscillators? Mail_Attachment -- Have Fun, Brooke Clarke http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Charles Steinmetz wrote: Karen wrote: Really I am looking for a source with a good short range stability performance between 1 and 10 seconds. I concur with everything Tom wrote. If that is your goal, the best you are likely to do is with a good, free-running (non-disciplined) OCXO. That will have the best stability you can get (short of a hydrogen maser) from 0.1 seconds to ~100 seconds. For the convenience of having automatic drift correction, many of us use GPSDOs and turn off the disciplining before making measurements. I might consider a PRS10 as an alternative, but certainly no other compact Rb, disciplined or otherwise. Even the rack-mount HP 5065 can't beat a good free-running OCXO. Many of the compact Rb units (and especially the FE5680) have nasty output waveforms with many spurs. 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. ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Many thanks for all your recommendations. Let me provide more details for understanding of my task. I am playing with a GPSDO project on base of uBlox NEO-7M (http://www.ra3apw.ru/ublox-neo-7m-ocxo-gpsdo/) - sorry, text in Russian. One of the main step – ADEV measurement of a developed GPSDO. My ADEV measure stand consists of a frequency counter Pendulum CNT-91 with TimeBase option 19 + GPIB interface + KE5FX TimeLab software (TNX again John). As option for CNT-91’s reference source I can use a homemade GPSDO on base of G3RUH design. IMHO, in this condition a frequency stability of my GPSDO project should be higher than a stability of CNT-91’s reference OCXO. Taking into account that rubidium source has a better short range stability than OCXO or GPSDO I hope to find an external rubidium as 10 MHz reference source instead of internal OCXO of counter. If my reasoning is not right could you please correct them as I am not an expert in this area. Karen, ra3apw Date: Sun, 26 Oct 2014 14:45:00 -0700 From: Brooke Clarke bro...@pacific.net Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen wrote: Really I am looking for a source with a good short range stability performance between 1 and 10 seconds. I concur with everything Tom wrote. If that is your goal, the best you are likely to do is with a good, free-running (non-disciplined) OCXO. That will have the best stability you can get (short of a hydrogen maser) from 0.1 seconds to ~100 seconds. For the convenience of having automatic drift correction, many of us use GPSDOs and turn off the disciplining before making measurements. I might consider a PRS10 as an alternative, but certainly no other compact Rb, disciplined or otherwise. Even the rack-mount HP 5065 can't beat a good free-running OCXO. Many of the compact Rb units (and especially the FE5680) have nasty output waveforms with many spurs. 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] 10MHz Rubidium reference source for frequency counter
Karen, a good double oven OCXO such as Morion etc will give you much better short term stability than most Rudidiums can give you. By a factor of 10 or even 100 sometimes below 10s measurement interval. The Rb's are better anywhere from 100s to many 1000 seconds. bye, Said In a message dated 10/27/2014 13:57:14 Pacific Daylight Time, ra3...@mail.ru writes: Many thanks for all your recommendations. Let me provide more details for understanding of my task. I am playing with a GPSDO project on base of uBlox NEO-7M (http://www.ra3apw.ru/ublox-neo-7m-ocxo-gpsdo/) - sorry, text in Russian. One of the main step – ADEV measurement of a developed GPSDO. My ADEV measure stand consists of a frequency counter Pendulum CNT-91 with TimeBase option 19 + GPIB interface + KE5FX TimeLab software (TNX again John). As option for CNT-91’s reference source I can use a homemade GPSDO on base of G3RUH design. IMHO, in this condition a frequency stability of my GPSDO project should be higher than a stability of CNT-91’s reference OCXO. Taking into account that rubidium source has a better short range stability than OCXO or GPSDO I hope to find an external rubidium as 10 MHz reference source instead of internal OCXO of counter. If my reasoning is not right could you please correct them as I am not an expert in this area. Karen, ra3apw Date: Sun, 26 Oct 2014 14:45:00 -0700 From: Brooke Clarke bro...@pacific.net Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hello Karen, I think you are confusing and/or mixing the terms stability and accuracy with respect to your project. It all depends upon your measurement period and the property being measured. A GPSDO will never beat a truly, very high quality OCXO on a short term basis in the stability department. For that matter most Rubidium's that you find on auction sites won't either. Accuracy, on the other hand, is another matter. A good and properly setup GPSDO will provide long term accuracy (years) with a Rubidium (months), calibrated, in second place and the very high quality OCXO (up to days) is last, again, with respect to accuracy. BillWB6BNQ Karen Tadevosyan wrote: Many thanks for all your recommendations. Let me provide more details for understanding of my task. I am playing with a GPSDO project on base of uBlox NEO-7M (http://www.ra3apw.ru/ublox-neo-7m-ocxo-gpsdo/) - sorry, text in Russian. One of the main step – ADEV measurement of a developed GPSDO. My ADEV measure stand consists of a frequency counter Pendulum CNT-91 with TimeBase option 19 + GPIB interface + KE5FX TimeLab software (TNX again John). As option for CNT-91’s reference source I can use a homemade GPSDO on base of G3RUH design. IMHO, in this condition a frequency stability of my GPSDO project should be higher than a stability of CNT-91’s reference OCXO. Taking into account that rubidium source has a better short range stability than OCXO or GPSDO I hope to find an external rubidium as 10 MHz reference source instead of internal OCXO of counter. If my reasoning is not right could you please correct them as I am not an expert in this area. Karen, ra3apw Date: Sun, 26 Oct 2014 14:45:00 -0700 From: Brooke Clarke bro...@pacific.net Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen, PRS-10.. Sent From iPhone On Oct 26, 2014, at 10:46, Karen Tadevosyan ra3...@mail.ru wrote: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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] 10MHz Rubidium reference source for frequency counter
I know what your next question will be, I see that these rubidium oscillators all can be adjusted over a range that falls on both sides of 10MHz. What should I use as a calibration source to adjust my rubidium oscillator? Then you think If I have this 10MHz calibration source, why not just run the frequency counter off that and not bother with the rubidium? OK there is good reason to use Rb. It warms up fast, it remains stable for a long time. It is self contained and compact and does not rely on any external signals. On Sun, Oct 26, 2014 at 10:46 AM, Karen Tadevosyan ra3...@mail.ru wrote: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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. -- Chris Albertson Redondo Beach, California ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Karen To a large degree it depends on what you want to use it for. HP 5065 is considered top of the line and PRS 10 is a very nice Rb but lately I have seen lamp oscillator failures. I like FRK, well documented easy to modify and if done right super performance. As part of a GPSDO project using the FE 5680A that my Swiss friend Juerg and I have worked on for a year and is now in Beta test is also an alternative. We are done with FE 5680A so Juerg has his along with a GPSDO for sale. If you want a affordable Rb that has GPS control that is an alternative. Unit has less than 6 month running time so like new. Bert Kehren in Miami In a message dated 10/26/2014 2:01:17 P.M. Eastern Daylight Time, ra3...@mail.ru writes: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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] 10MHz Rubidium reference source for frequency counter
Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. You might take a look at the plots at http://www.ke5fx.com/rb.htm if you haven't already, and also at http://febo.com/pages/oscillators/rubes/ . These will give you an idea what to expect from some of the popular Rb models. The PRS10 mentioned by Said is a great choice for long-term measurements, but it's also the most expensive of the ones you are likely to find on the surplus market. Interestingly, if you want the best performance between 1 and 10 seconds the PRS10 is actually the worst of the ones I measured. But the CNT-91 probably will not be able to see the difference, since its single-shot resolution is a bit worse than the HP 5370B counter (see attached plot). Caveats: 1) No rubidium standard will give its best long-term performance until it has been allowed to run undisturbed for at least a few days. 2) These plots shouldn't be taken too seriously because of the (very) small number of examples tested under inconsistent environmental conditions. -- 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] 10MHz Rubidium reference source for frequency counter
Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Karen Tadevosyan wrote: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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] 10MHz Rubidium reference source for frequency counter
Brooke, One use for the R variants of the Fluke/Pendulum counters is/was for calibrating base-stations. They had issues with ovens and turning the counter to the side as you lifted it up. A rubidium inside solved that in a nice way. It's not all government work you know. :) Cheers, Magnus On 10/26/2014 10:45 PM, Brooke Clarke wrote: Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Karen Tadevosyan wrote: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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. ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
I know what your next question will be, I see that these rubidium oscillators all can be adjusted over a range that falls on both sides of 10MHz. What should I use as a calibration source to adjust my rubidium oscillator? Then you think If I have this 10MHz calibration source, why not just run the frequency counter off that and not bother with the rubidium? OK there is good reason to use Rb. It warms up fast, it remains stable for a long time. It is self contained and compact and does not rely on any external signals. You really have to tweak the heck out of a GPSDO to achieve medium-term stability comparable to that of a cheap rubidium (see attached comparison against typical and exceptional GPSDOs.) Of course, at taus greater than several hours the GPSDO will eventually win. -- 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] 10MHz Rubidium reference source for frequency counter
Hi Magnus: I've also heard that in order to calibrate rack mount crystal oscillators in instruments they need to be in the same orientation as when mounted in the rack. So you can not remove the instrument from the rack and turn it on it's side for the cal. So for some instruments that means mounting them in an empty rack and laying on your back like working under a car. Mail_Attachment -- Have Fun, Brooke Clarke http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Magnus Danielson wrote: Brooke, One use for the R variants of the Fluke/Pendulum counters is/was for calibrating base-stations. They had issues with ovens and turning the counter to the side as you lifted it up. A rubidium inside solved that in a nice way. It's not all government work you know. :) Cheers, Magnus On 10/26/2014 10:45 PM, Brooke Clarke wrote: Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Karen Tadevosyan wrote: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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. ___ time-nuts mailing 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] 10MHz Rubidium reference source for frequency counter
Hi Brooke, I wonder which instruments that would be, as most of them calibrate easily standing flat on the bench. Cheers, Magnus On 10/26/2014 11:01 PM, Brooke Clarke wrote: Hi Magnus: I've also heard that in order to calibrate rack mount crystal oscillators in instruments they need to be in the same orientation as when mounted in the rack. So you can not remove the instrument from the rack and turn it on it's side for the cal. So for some instruments that means mounting them in an empty rack and laying on your back like working under a car. Mail_Attachment -- Have Fun, Brooke Clarke http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Magnus Danielson wrote: Brooke, One use for the R variants of the Fluke/Pendulum counters is/was for calibrating base-stations. They had issues with ovens and turning the counter to the side as you lifted it up. A rubidium inside solved that in a nice way. It's not all government work you know. :) Cheers, Magnus On 10/26/2014 10:45 PM, Brooke Clarke wrote: Hi Karen: The ones you mention are all stand alone Rb oscillators that need to be calibrated to set their frequency. This was the historical way that crystal oscillators were calibrated every year or so. The great advantage of Rb over crystal oscillators is that their drift is specified in months instead of days. A much better - more modern idea - is the GPS Disciplined Oscillator (GPSDO). It keeps the oscillator calibrated in real time. A popular crystal based GPSDO is the Trimble ThunderBolt: http://www.prc68.com/I/ThunderBolt.shtml Another crystal based GPSDO is the HP Z3805: http://www.prc68.com/I/Z3805A.html There are many more commercial GPSDOs and this list has discussions that show they can be a do it yourself project for under maybe $10, but require a number of sophisticated skills. I have the just released LTE-Lite GPSDO Evaluation Kit with 10MHz TCXO on order. Seems to offer good performance for the dollar. http://www.ebay.com/itm/171504585820 The only advantage of a Rb GPSDO over a crystal GPSDO is for the case where the GPS updating has not happened for some time. This might be due to a power failure lasting some days or that the oscillator will be used where there's no GPS access and it only gets calibrated then used much later. The Stanford Research PRS-10 Rb oscillator can be used stand alone where it time stamps an external 1 Pulse Per Second input, or as part of a GPSDO where an external GPS receiver supplies it with a 1 PPS input. http://www.prc68.com/I/PRS10.shtml The Thunderbolt can be custom modified to drive an external Rb oscillator, like the ones you mentioned, but that requires some technical sophistication. Note the ThunderBolt and Z3805 are complete GPSDOs in a box, just connect power and a GPS antenna. The PRS-10 requires an external GPS receiver and antenna. A a practical matter that means it's more work to maintain the PRS-10 because there's more opportunity for problems like disconnecting a cable. PS Stanford Research offered a version of their SR620 Time Interval counter that included a Rb oscillator (not a GPSDO) that some government agencies purchased, but for normal use you really don't need a Rb oscillator, so the CNT-91R appears to be a similar way so sell it to a government with a lot of money to spare. So don't feel pressured to use an Rb oscillator. http://www.prc68.com/I/TandFTE.shtml#SR620 Mail_Attachment -- Have Fun, Brooke Clarke, N6GCE http://www.PRC68.com http://www.end2partygovernment.com/2012Issues.html http://www.prc68.com/I/DietNutrition.html Karen Tadevosyan wrote: Hello All, Can I have your recommendation regarding a choice of 10 MHz rubidium source (available now on eBay like FE-5680; LPRO-101; LPFRS; FRS etc.) as a reference signal for my frequency counter Pendulum CNT-91. Thanks in advance. Karen, ra3apw --- Это сообщение свободно от вирусов и вредоносного ПО благодаря защите от вирусов 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. ___ time-nuts mailing 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
