Re: [time-nuts] GPS receiver testing
Hi all, I'm new here (been reading for a while, but this is my first reply) so sorry if I'm not doing this reply thing right. There is a Spirent simulator down the hall from me that I tested the output PPS on a few months ago using an Agilent 53230A (OCXO option). While I don't remember the exact numbers, I believe the time interval between each successive PPS was good to within a few tens of picoseconds. I'd be happy to test it again if people were interested in seeing more detailed results. Magnus, Most GPS simulators only have an OCXO in them which are not very good for making precision timing measurements. The simulator needs to have a rubidium in it or have an external input for a more precision frequency source. Regards,Doug Connected by DROID on Verizon Wireless ___ 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 receiver testing
Thank you to all of the respondents. There is a lot of great information in the replies. Based on more interaction with our customer it's looking like we may need to send the unit(s) out to a facility that specializes in this area. Best regards, John -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Jim Lux Sent: Wednesday, October 31, 2012 8:39 PM To: time-nuts@febo.com Subject: Re: [time-nuts] GPS receiver testing Hello Said, I'm familiar with your postings on the Time Nuts list, so I thought I'd ask your advice. I searched the Time Nuts archive, but didn't come up with what I was looking for (reference to a good GPS tutotial). We have a GPS project in-house that requires us to characterize receiver module performance. We have a Litepoint IQ-Nav box with several stored scenarios, but no other signal generators with GPS personalities in them. We need to measure position accuracy and time accuracy. We may also need to get some characterization of the 1 PPS output. There's folks at JPL who do this kind of thing all the time.. I'll ask who you should contact.. It kind of matters on what you're testing for, but I'm sure there's someone who can point you in the right direction. Jim ___ 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 receiver testing
John, thanks for your email, I am replying to Time Nuts as well as there is a lot of knowledgeable folks here that can help. In terms of a GPSDO tutorial, take a look at the HP papers linked on the JLT website under the Links Of Interest and Related Whitepaper sections: _http://www.jackson-labs.com/index.php/support_ (http://www.jackson-labs.com/index.php/support) In terms of how to measure the 1PPS accuracy and how to set up the equipment, see the paper labeled Critical evaluation of the Motorola M12+ receiver on that page. Explains how to set up the Agilent counter I mention below. There is a lot of threads in the time nuts archives discussing the pro's and con's of different equipment, but let me give you a quick guide to what worked well for me: For 1PPS measurements and frequency stability down to the ~2E-010 level per second you can get a low-cost Agilent 53132A counter. In order to check GPS position accuracy, you may want to get a timing GPS receiver that can do position-averaging using Auto Survey features. Such as the Trimble Thunderbolts, the JLT Mini-JLT GPSDO, or the HP58503A units. The JLT unit is the only unit using WAAS augmentation, so it probably has a much quicker and more accurate position indication than the other units mentioned. Let it average the position of your antenna for a day or two, and you will likely have an accuracy at the centimeter level (horizontal) and at the foot level vertical. Beware of different GPS datums, e.g. MSL versus GPS height indications etc. You can use those GPSDO as a reference for your counter as well. The above equipment can be had with a couple of days shipping time from Ebay, at around $1500 to $2000 total and will serve you well for a very long time, and the resolution of the counter (150ps) is likely much higher than the GPS sawtooth error from the GPS you mentioned. If you need much more accuracy and resolution, get a Wavecrest DTS time interval analyzer from Ebay for around $800, those have picosecond averaged noise floors, femtosecond hardware resolution, and 10ps single shot resolution. In order to measure the 1PPS stability and accuracy, you would input the GPSDO reference 1PPS and your GPS 1PPS into the counter, and set it to T1 to T2 time interval measurement can capture that data. You may or may not use an external 10MHz reference for the counter doing this measurement, it shouldn't make a difference to your results.Then download Ulrich Bangerts excellent freeware Plotter program to do the time-stability analysis (search Google for Bangert Plotter and you will find his website). Please note that you may or may not want to use the GPS receiver sawtooth correction data on your dataset, either manually (using Excel to subtract the offset error), via a delay line, or other mechanism in your system. It will make a significant difference in your stability. To measure frequency, feed the GPSDO reference into the ref-in port of the counter, your DUT to the A input, then use the offset feature to remove the carrier frequency, then capture the frequency offset of your source to a file, and again use Ulrich's plotter to give you the time-stability info etc. Be warned, once you start on above path, you are likely never to stop searching for the holy grail of references, and measurement equipment.. Hope that helps, Said In a message dated 10/31/2012 12:36:55 Pacific Daylight Time, jlofg...@lsr.com writes: Hello Said, I'm familiar with your postings on the Time Nuts list, so I thought I'd ask your advice. I searched the Time Nuts archive, but didn't come up with what I was looking for (reference to a good GPS tutotial). We have a GPS project in-house that requires us to characterize receiver module performance. We have a Litepoint IQ-Nav box with several stored scenarios, but no other signal generators with GPS personalities in them. We need to measure position accuracy and time accuracy. We may also need to get some characterization of the 1 PPS output. I know that you do these types of measurements frequently. Could you point me to a good reference on correctly performing these tests and, maybe, describe the equipment you are using? We're checking through Agilent and RS application papers, but you seem to have a lot of the required knowledge readily available. Any help you could provide would be appreciated. Best regards, John ___ 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 receiver testing
Thank you very much for the thorough reply. Best Regards, John From: saidj...@aol.com [mailto:saidj...@aol.com] Sent: Wednesday, October 31, 2012 3:17 PM To: John Lofgren; time-nuts@febo.com Subject: Re: GPS receiver testing John, thanks for your email, I am replying to Time Nuts as well as there is a lot of knowledgeable folks here that can help. In terms of a GPSDO tutorial, take a look at the HP papers linked on the JLT website under the Links Of Interest and Related Whitepaper sections: http://www.jackson-labs.com/index.php/support In terms of how to measure the 1PPS accuracy and how to set up the equipment, see the paper labeled Critical evaluation of the Motorola M12+ receiver on that page. Explains how to set up the Agilent counter I mention below. There is a lot of threads in the time nuts archives discussing the pro's and con's of different equipment, but let me give you a quick guide to what worked well for me: For 1PPS measurements and frequency stability down to the ~2E-010 level per second you can get a low-costAgilent 53132A counter. In order to check GPS position accuracy, you may want to get a timing GPS receiver that can do position-averaging using Auto Survey features. Such as the Trimble Thunderbolts, the JLT Mini-JLT GPSDO, or the HP58503A units. The JLT unit is the only unit using WAAS augmentation, so it probably has a much quicker and more accurate position indication than the other units mentioned. Let it average the position of your antenna for a day or two, and you will likely have an accuracy at the centimeter level (horizontal) and at the foot level vertical. Beware of different GPS datums, e.g. MSL versus GPS height indications etc. You can use those GPSDO as a reference for your counter as well. The above equipment can be had with a couple of days shipping time from Ebay, at around $1500 to $2000 totaland will serve you well for a very long time, and the resolution of the counter (150ps) is likely much higher than the GPS sawtooth error from the GPS you mentioned. If you need much more accuracy and resolution, get a Wavecrest DTS time interval analyzer from Ebay for around $800, those have picosecond averaged noise floors, femtosecond hardware resolution, and 10ps single shot resolution. In order to measure the 1PPS stability and accuracy, you would input the GPSDO reference 1PPS and your GPS 1PPS into the counter, and set it to T1 to T2 time interval measurement can capture that data. You may or may not use an external 10MHz reference for the counter doing this measurement, it shouldn'tmake a difference to your results.Then download Ulrich Bangerts excellent freeware Plotter program to do the time-stability analysis (search Google for Bangert Plotter and you will find his website). Please note that you may or may not want to use the GPS receiver sawtooth correction data on your dataset, either manually (using Excel to subtract the offset error), viaa delay line, or other mechanism in your system. It will make a significant difference in your stability. To measure frequency, feed the GPSDO reference into the ref-in port of the counter, your DUTto the A input, then use the offset feature to remove the carrier frequency, then capture the frequency offset of your source to a file, and again use Ulrich's plotter to give you the time-stability info etc. Be warned, once you start on above path, you are likely never to stop searching for the holy grailof references, and measurement equipment.. Hope that helps, Said In a message dated 10/31/2012 12:36:55 Pacific Daylight Time, jlofg...@lsr.com writes: Hello Said, I'm familiar with your postings on the Time Nuts list, so I thought I'd ask your advice. I searched the Time Nuts archive, but didn't come up with what I was looking for (reference to a good GPS tutotial). We have a GPS project in-house that requires us to characterize receiver module performance. We have a Litepoint IQ-Nav box with several stored scenarios, but no other signal generators with GPS personalities in them. We need to measure position accuracy and time accuracy. We may also need to get some characterization of the 1 PPS output. I know that you do these types of measurements frequently. Could you point me to a good reference on correctly performing these tests and, maybe, describe the equipment you are using? We're checking through Agilent and RS application papers, but you seem to have a lot of the required knowledge readily available. Any help you could provide would be appreciated. Best regards, John ___ 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 receiver testing
John, I would advice you to look at what Spirent and Pendulum can offer you with regards to GPS simulators. The Pendulum GPS simulation should be of interest as it is a fairly decent cost, where as the Spirent equipment is really really good. One way to characterize PPS properties is to do it against a GPS simulator, which would allow you to remove some uncertainties compared to live signal. Said's pointers are indeed a good start, so I only wanted to give an extra dimension which you might need. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS receiver testing
Magnus, Most GPS simulators only have an OCXO in them which are not very good for making precision timing measurements. The simulator needs to have a rubidium in it or have an external input for a more precision frequency source. Regards,Doug Connected by DROID on Verizon Wireless -Original message- From: Magnus Danielson mag...@rubidium.dyndns.org To: time-nuts@febo.com Sent: Thu, Nov 1, 2012 00:11:51 GMT+00:00 Subject: Re: [time-nuts] GPS receiver testing John, I would advice you to look at what Spirent and Pendulum can offer you with regards to GPS simulators. The Pendulum GPS simulation should be of interest as it is a fairly decent cost, where as the Spirent equipment is really really good. One way to characterize PPS properties is to do it against a GPS simulator, which would allow you to remove some uncertainties compared to live signal. Said's pointers are indeed a good start, so I only wanted to give an extra dimension which you might need. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS receiver testing
When I had (for a very limited time) the SMBV100A from RS loaded with the GPS simulator, I connected it to the 10MHz and PPS output of my GPSDO. Unfortunately there was not time enough to take measures... anyway the PPS output of the GPS under test aligned with the incoming PPS. On Thu, Nov 1, 2012 at 1:31 AM, k4...@aol.com k4...@aol.com wrote: Magnus, Most GPS simulators only have an OCXO in them which are not very good for making precision timing measurements. The simulator needs to have a rubidium in it or have an external input for a more precision frequency source. Regards,Doug Connected by DROID on Verizon Wireless -Original message- From: Magnus Danielson mag...@rubidium.dyndns.org To: time-nuts@febo.com Sent: Thu, Nov 1, 2012 00:11:51 GMT+00:00 Subject: Re: [time-nuts] GPS receiver testing John, I would advice you to look at what Spirent and Pendulum can offer youwith regards to GPS simulators. The Pendulum GPS simulation should be ofinterest as it is a fairly decent cost, where as the Spirent equipmentis really really good. One way to characterize PPS properties is to doit against a GPS simulator, which would allow you to remove someuncertainties compared to live signal. Said's pointers are indeed a good start, so I only wanted to give anextra dimension which you might need. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ 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 receiver testing
On 11/01/2012 01:31 AM, k4...@aol.com wrote: Magnus, Most GPS simulators only have an OCXO in them which are not very good for making precision timing measurements. The simulator needs to have a rubidium in it or have an external input for a more precision frequency source. Not really, the GPS simulator only needs to be more stable than the GPS receiver it is testing, as short-term differences is measured. Adding a rubidium should be trivial if it would be needed. Cheers, Magnus ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS receiver testing
Measuring 1PPS stability requires a counter and a reference. As Said mentioned, the counter requirement is fairly simple: any sub-nanosecond counter like a 53132A is fine (higher resolution is wasted anyway). But the reference requirement is the more difficult. If you want to measure down to nanoseconds a day, you need either a good cesium or an H-maser. That's why most people send their GPSDO to a timing lab for these measurements. If you use a GPSDO as a reference you find yourself in the awkward position of using one GPS product to measure another GPS product. I suspect this would create more questions than answers. Alternatively if you use a GPS simulator you can cancel out the instability of the reference. In theory. If you're just trying to measure down to the 10 ns or 100 ns level, then any old Rb or Cs reference will do. John is also interested in measuring *accuracy*. This is really hard since it involves bringing equipment to NIST or USNO or similar timing installation and/or using antenna test chambers, etc. Please read John Plumb's paper: http://tycho.usno.navy.mil/ptti/ptti2002/paper29.pdf and Rick Hambly's paper: http://tycho.usno.navy.mil/ptti/ptti2002/paper9.pdf For further information, google for: gps absolute calibration and you'll find many papers. All the UTC(k) need to do this for their submissions to BIPM so there is no shortage of information on this topic. /tvb ___ 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 receiver testing
Hello Said, I'm familiar with your postings on the Time Nuts list, so I thought I'd ask your advice. I searched the Time Nuts archive, but didn't come up with what I was looking for (reference to a good GPS tutotial). We have a GPS project in-house that requires us to characterize receiver module performance. We have a Litepoint IQ-Nav box with several stored scenarios, but no other signal generators with GPS personalities in them. We need to measure position accuracy and time accuracy. We may also need to get some characterization of the 1 PPS output. There's folks at JPL who do this kind of thing all the time.. I'll ask who you should contact.. It kind of matters on what you're testing for, but I'm sure there's someone who can point you in the right direction. Jim ___ 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 receiver testing
On 10/31/12 6:17 PM, Tom Van Baak wrote: Please read John Plumb's paper: http://tycho.usno.navy.mil/ptti/ptti2002/paper29.pdf and Rick Hambly's paper: http://tycho.usno.navy.mil/ptti/ptti2002/paper9.pdf At JPL, most of the calibration and performance testing and such is done using an antenna outside with actual signals, rather than a test set (although we do have some fancy signal simulators). Depending on what data you can get out of your receiver, working with the GIPSY offline processing system might be useful. IN a post processing sense, you can determine what your receiver *should* have been returning. This morning, I was in a meeting where the JPL GPS folks were talking about improving the terrestrial reference frame accuracy from cm scale to mm scale over the next few years (and tieing it to celestial references as well). This has to account for all the things like solid earth tides, plate movement, etc. ___ 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.