Re: [time-nuts] GPS 1PPS ultimate accuracy
Hi Andrea, Making measurements of quartz oscillator aging is much easier than you think and requires minimal equipment. In particular all you need is a GPS 1PPS and a simple counter. No need to worry about sawtooth. Any $20 GPS/1PPS receiver will work. Consider this rough example of measuring for one week a OCXO with 1e-10/day frequency drift rate. Say on day 1 it is 1e-10 low in frequency. It will lose 0.1 ns per second. That's a very small amount and expensive to measure. But who cares. You're not trying to measure phase noise, or short-term stability, or frequency; all you're trying to measure is frequency drift. So let it run all day. By the end of the day those 0.1 ns have added up to 8.6 us. One data point. 8 microseconds is a lot. You can measure this with a $1 PIC or a $10 Arduinio. Use either time interval or timestamping methods. See PIC examples at www.leapsecond.com/pic/ although you can turn just about any microcontroller into a microsecond, or sub-microsecond counter and results over serial or USB to a PC for logging. On day 2 it has drifted to be 2e-10 low in frequency, so it will lose an additional 17.2 us of time, etc. Another data point. By the end of the week, the slowly aging oscillator is 7e-10 low in frequency and will lose 60 us that day. This simple experiment would give you 7 data points which would nicely show your oscillator drift rate. You could collect data more than once a day if you wanted, like every hour or every minute. Differentiate the time error to get frequency error. Differentiate frequency to get frequency drift rate. Or just do a quadratic fit of the raw time error data. It requires so little hardware that you could easily let it run for a month, or year and collect wonderful data. The more the oscillator drifts the larger the time measurements are so the easier they are to measure with accuracy. This setup might even work for Rubidium. On the one hand Rubidium drift rates are 10x to 100x less than OCXO so your times will not grow nearly as rapidly, making precise measurements more difficult. On the other hand, Rubidium drift rates are so low that you would want to measure for months instead of weeks. In the end the two factors may balance themselves. So I don't think you need nanosecond counters or fancy sawtooth corrected GPS timing receivers or GPSDO or measurements every second. A slowly aging oscillator is very easy to measure, mostly because, in order to measure aging you need many days or weeks or months of data. The longer the measurement time, the less it matters what the resolution of the counter (or the GPS 1PPS) is or how quickly you collect data. /tvb - Original Message - From: Andrea Baldoni erm1ea...@ermione.com To: time-nuts@febo.com Sent: Monday, January 12, 2015 2:59 AM Subject: [time-nuts] GPS 1PPS ultimate accuracy Hello all. I am planning to do some experiments to evaluate the aging of oscillators (this one of the reasons I'm willing to buy the Milleren without EFC). What I would like to do exactly is to sample the total of a counter (of suitable number of bits, taking in account the fact that it will overflow) whose clock is the DUT. The sampling interval could come from a (long time based on a) sawtooth uncorrected PPS from a cheap GPS, a sawtooth corrected from a good one (perhaps the Lucent GPSDO), or a computer using NTP. Each of these sources should reach a goal stability (say, 1 part in 10^13) after averaging them on a different (and very high I suppose) number of seconds (averaging them for an infinity number of seconds should give the stability of the underlying reference clock, but I'm willing to stop sooner...). I know there's no reason to go 1E-13 when the Milliren couldn't go that far, but the DUT may be also something else like a FE-5680A). The sawtooth uncorrected GPS receiver may never yeld a good stability in the short term, but in the long one it should as well because the internal clock jitter would average results. If I'm using the correct teminology, after what tau the ADEV graph of the different references intersect the 1E-13? By the way, the stability of the TAI is known or, because it's the reference one, it has zero deviation for definition (so you can reach its ultimate stability through GPS really only at the infinity...)? Best regards, Andrea Baldoni ___ 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] GPS 1PPS ultimate accuracy
On Mon, Jan 12, 2015 at 06:16:01PM -0500, Bob Camp wrote: Actually it’s a bit worse than you might expect. The uncorrected sawtooth will give you about 20 ns of wander. At the one day level, GPS without some sort of ionosphere help (like a dual frequency receiver) will add another 10 ns or so to that. Net, your pps is spread over a 30 ns range. Hello Bob. Thank you, now I have a better idea. I understand that the NTP is completely ruled out and also between GPS there is a strong difference. With things like 5335’s running around for cheap prices, I would suggest doing this with a counter. You are going to spend a lot of days getting very much data. Your time’s got to be worth something …. Actually I own a Racal 1995 that should be better than the 5335 with its 1ns single shot resolution. However, I don't still own a GPSDO to reference the counter so how do you suggest to use it? I should use total A over B with the DUT in A and the PPS in B? Best regards, Andrea Baldoni ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS 1PPS ultimate accuracy
On Mon, Jan 12, 2015 at 03:10:39PM +0100, Attila Kinali wrote: The GNSS Timing AppNote for the LEA6-T receiver[1] will give you an idea what jitter you get with GPS. Please be aware that these measurements were done with an antenna located at a _good_ position (ontop of a 4 story building with no other high buildings around). Unless you have a simlarly good location you will have worse performance. Ciao Attila. By the way, I see there are LEA-6T from Hong Kong at 49 USD shipping included. If those are not a fake and I can extract the PPS from them, do you suggest this as the best GPS for the price actually available for timing? It would average out if and only if the sawtooth correction would be completely independent of anything else. But it isn't. This results in effects where the cycle to cycle jitter is quite low, but there is a large offset in the sawtooth correction. This is know as hanging bridges in the GNSS world. I can use the sawtooth correction with LEA-6T but if I am using it with a normal TIC I should obtain a way either to apply the correction in hardware, or to capture the numbers and postprocess them together with data from the TIC. Probably the simple solution is a GPSDO where everything is already done? Best regards, Andrea Baldoni ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS 1PPS ultimate accuracy
Hi On Jan 15, 2015, at 4:56 AM, Andrea Baldoni erm1ea...@ermione.com wrote: On Mon, Jan 12, 2015 at 06:16:01PM -0500, Bob Camp wrote: Actually it’s a bit worse than you might expect. The uncorrected sawtooth will give you about 20 ns of wander. At the one day level, GPS without some sort of ionosphere help (like a dual frequency receiver) will add another 10 ns or so to that. Net, your pps is spread over a 30 ns range. Hello Bob. Thank you, now I have a better idea. I understand that the NTP is completely ruled out and also between GPS there is a strong difference. With things like 5335’s running around for cheap prices, I would suggest doing this with a counter. You are going to spend a lot of days getting very much data. Your time’s got to be worth something …. Actually I own a Racal 1995 that should be better than the 5335 with its 1ns single shot resolution. However, I don't still own a GPSDO to reference the counter so how do you suggest to use it? Divide both of the things you are testing down to 1 pps. Trigger the start on one and the stop on the other. Read out the difference to the 1 or 2 ns resolution of the counter. That’s going to be ~ 100 X better than measurement with the rolling counter. Bob I should use total A over B with the DUT in A and the PPS in B? Best regards, Andrea Baldoni ___ 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] GPS 1PPS ultimate accuracy
Hello all. I am planning to do some experiments to evaluate the aging of oscillators (this one of the reasons I'm willing to buy the Milleren without EFC). What I would like to do exactly is to sample the total of a counter (of suitable number of bits, taking in account the fact that it will overflow) whose clock is the DUT. The sampling interval could come from a (long time based on a) sawtooth uncorrected PPS from a cheap GPS, a sawtooth corrected from a good one (perhaps the Lucent GPSDO), or a computer using NTP. Each of these sources should reach a goal stability (say, 1 part in 10^13) after averaging them on a different (and very high I suppose) number of seconds (averaging them for an infinity number of seconds should give the stability of the underlying reference clock, but I'm willing to stop sooner...). I know there's no reason to go 1E-13 when the Milliren couldn't go that far, but the DUT may be also something else like a FE-5680A). The sawtooth uncorrected GPS receiver may never yeld a good stability in the short term, but in the long one it should as well because the internal clock jitter would average results. If I'm using the correct teminology, after what tau the ADEV graph of the different references intersect the 1E-13? By the way, the stability of the TAI is known or, because it's the reference one, it has zero deviation for definition (so you can reach its ultimate stability through GPS really only at the infinity...)? Best regards, Andrea Baldoni ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS 1PPS ultimate accuracy
Hi Actually it’s a bit worse than you might expect. The uncorrected sawtooth will give you about 20 ns of wander. At the one day level, GPS without some sort of ionosphere help (like a dual frequency receiver) will add another 10 ns or so to that. Net, your pps is spread over a 30 ns range. The output of the OCXO is at 5 MHz, the Rb is at 10 MHz. Maybe you double the OCXO to 10 MHz. It only has a zero crossing every 100 ns. (200 ns if you don’t double it). You will have a 100 ns “dead zone” in your counter. That assumes it’s synchronous. If it’s a ripple counter, who knows what it will do. Net result, You have 30 ns of error, and a 100 ns resolution. Net is 130 ns. You will hit 1x10^-9 at a bit over 100 seconds. You will get to 1x10^-12 at around 13,000 seconds. Since it’s a dead zone, averaging really does not help you much. In fact, long averages will mess up the ADEV computations. If you have a goal that resolution should be 5X your data, then you get to 1x10^-12 data at around 80,000 seconds. For a good ADEV number, you would like about 100 samples. This gets you out to a 100 day run. Even for a minimalist number, you are running for 10 days. Any time you have a power interruption, the process re-starts. With things like 5335’s running around for cheap prices, I would suggest doing this with a counter. You are going to spend a lot of days getting very much data. Your time’s got to be worth something …. Bob On Jan 12, 2015, at 9:10 AM, Attila Kinali att...@kinali.ch wrote: Ciao Andrea, On Mon, 12 Jan 2015 11:59:26 +0100 Andrea Baldoni erm1ea...@ermione.com wrote: The sampling interval could come from a (long time based on a) sawtooth uncorrected PPS from a cheap GPS, a sawtooth corrected from a good one (perhaps the Lucent GPSDO), or a computer using NTP. The GNSS Timing AppNote for the LEA6-T receiver[1] will give you an idea what jitter you get with GPS. Please be aware that these measurements were done with an antenna located at a _good_ position (ontop of a 4 story building with no other high buildings around). Unless you have a simlarly good location you will have worse performance. Said Jackson reported some time ago that he got around 1us of jitter for a GPS receiver (i presume it was either a LEA5-T or a LEA6-T) behind a window. After he averaged the position for a long time (several days) manually and stored that in the receiver he got much better performance (sorry i cannot find the mail at the moment, you have to look for it in the archives yourself). NTP will give you a jitter in the range of 1-100ms, depending on your internet connection and its conguestion. On a local network based NTP system, you can expect jitter in the range of 10-100us IIRC. Each of these sources should reach a goal stability (say, 1 part in 10^13) after averaging them on a different (and very high I suppose) number of seconds (averaging them for an infinity number of seconds should give the stability of the underlying reference clock, but I'm willing to stop sooner...). I know there's no reason to go 1E-13 when the Milliren couldn't go that far, but the DUT may be also something else like a FE-5680A). To get to 1e-13 with GPS (assuming 1-10ns jitter) you need somewhere around 10k to 100k seconds. At these time scales, the temperature dependent deviation of your OCXO is likely to dominate your measurement. I would rather do a two step measurement. If you have a FE-5680A measure its drift with a tau in the 100ks-200ks region. Then use the FE-5680A as refrence to measure the drift of the OCXO in 10s-1000s timescales. If you do both continuously, you can apply some math and get out pretty good numbers (see three cornered hat method) Additional to GPS jitter you also have the deviation of GPS time in respect to TAI/UTC. This has been in recent years below 5ns (GPS vs UTC(USNO)). But because GPS time is steered to be close to UTC it will oscillate slightly around it. How much, i do not know. (But then the deviation between the different UTC realizations is larger) [2] The sawtooth uncorrected GPS receiver may never yeld a good stability in the short term, but in the long one it should as well because the internal clock jitter would average results. It would average out if and only if the sawtooth correction would be completely independent of anything else. But it isn't. This results in effects where the cycle to cycle jitter is quite low, but there is a large offset in the sawtooth correction. This is know as hanging bridges in the GNSS world. By the way, the stability of the TAI is known or, because it's the reference one, it has zero deviation for definition (so you can reach its ultimate stability through GPS really only at the infinity...)? There is an uncertainty number attached to TAI, but i dont know any numbers from the top of my head. I'm sure it is mentioned in the BIPM report somewhere.
Re: [time-nuts] GPS 1PPS ultimate accuracy
Ciao Andrea, On Mon, 12 Jan 2015 11:59:26 +0100 Andrea Baldoni erm1ea...@ermione.com wrote: The sampling interval could come from a (long time based on a) sawtooth uncorrected PPS from a cheap GPS, a sawtooth corrected from a good one (perhaps the Lucent GPSDO), or a computer using NTP. The GNSS Timing AppNote for the LEA6-T receiver[1] will give you an idea what jitter you get with GPS. Please be aware that these measurements were done with an antenna located at a _good_ position (ontop of a 4 story building with no other high buildings around). Unless you have a simlarly good location you will have worse performance. Said Jackson reported some time ago that he got around 1us of jitter for a GPS receiver (i presume it was either a LEA5-T or a LEA6-T) behind a window. After he averaged the position for a long time (several days) manually and stored that in the receiver he got much better performance (sorry i cannot find the mail at the moment, you have to look for it in the archives yourself). NTP will give you a jitter in the range of 1-100ms, depending on your internet connection and its conguestion. On a local network based NTP system, you can expect jitter in the range of 10-100us IIRC. Each of these sources should reach a goal stability (say, 1 part in 10^13) after averaging them on a different (and very high I suppose) number of seconds (averaging them for an infinity number of seconds should give the stability of the underlying reference clock, but I'm willing to stop sooner...). I know there's no reason to go 1E-13 when the Milliren couldn't go that far, but the DUT may be also something else like a FE-5680A). To get to 1e-13 with GPS (assuming 1-10ns jitter) you need somewhere around 10k to 100k seconds. At these time scales, the temperature dependent deviation of your OCXO is likely to dominate your measurement. I would rather do a two step measurement. If you have a FE-5680A measure its drift with a tau in the 100ks-200ks region. Then use the FE-5680A as refrence to measure the drift of the OCXO in 10s-1000s timescales. If you do both continuously, you can apply some math and get out pretty good numbers (see three cornered hat method) Additional to GPS jitter you also have the deviation of GPS time in respect to TAI/UTC. This has been in recent years below 5ns (GPS vs UTC(USNO)). But because GPS time is steered to be close to UTC it will oscillate slightly around it. How much, i do not know. (But then the deviation between the different UTC realizations is larger) [2] The sawtooth uncorrected GPS receiver may never yeld a good stability in the short term, but in the long one it should as well because the internal clock jitter would average results. It would average out if and only if the sawtooth correction would be completely independent of anything else. But it isn't. This results in effects where the cycle to cycle jitter is quite low, but there is a large offset in the sawtooth correction. This is know as hanging bridges in the GNSS world. By the way, the stability of the TAI is known or, because it's the reference one, it has zero deviation for definition (so you can reach its ultimate stability through GPS really only at the infinity...)? There is an uncertainty number attached to TAI, but i dont know any numbers from the top of my head. I'm sure it is mentioned in the BIPM report somewhere. Attila Kinali [1] http://www.u-blox.com/images/downloads/Product_Docs/Timing_AppNote_%28GPS.G6-X-11007%29.pdf [2] GPS time and its steering to UTC(USNO), presentatin by Edward Powers, 2009 http://www.gps.gov/multimedia/presentations/2009/09/ICG/powers.pdf -- It is upon moral qualities that a society is ultimately founded. All the prosperity and technological sophistication in the world is of no use without that foundation. -- Miss Matheson, The Diamond Age, Neil Stephenson ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.