Hi BobS,

Thanks for sharing that. To follow-up on recent emails...

When you start to push the limits of your own test equipment it's good to employ tricks such as 3-corner hat in order to get a few dB better measurement. But there's a limit to that. And you probably don't want me to suggest that you run 3 GPSDO, 3 counters and 3 cesiums all simultaneously.

I've attached two plots showing 4+ days of your GPSDO data and 4+ days of a TBolt (factory defaults, untuned)

bob-gpsdo-3.gif -- Notice how much your GPSDO (blue trace) wanders all over the place.

bob-gpsdo-5.gif -- ADEV comparison. See below for comments.

When I look at the ADEV for your data set it's pretty clear that the limiting factor on the left (short tau) is the resolution of your 5370A counter. That's why the TBolt looks better. Don't worry about that.

In the middle of the ADEV plot it would appear that both your 5370A counter and your PRS-45A cesium reference are sufficiently good to properly measure your GPSDO. That's good. The TBolt looks bad in the middle but that's simply because it was using a way-too-short default time constant. With proper tuning the red trace would be mostly flat out to 1000 seconds.

http://leapsecond.com/pages/tbolt-8d/sigma1.gif from http://www.leapsecond.com/pages/tbolt-8d/
you'll note the sidereal "blip" is below 4e-14. Your ADEV doesn't get near that level.

I say this because a rule-of-thumb is that a good M12+T timing receiver should get you down to 3 ns per day RMS. And a good ublox-6T can get you to 3 or 2 ns per day RMS. Your data shows something more like 5 ns -- which for a GPSDO is worse than no GPSDO at all. So something's wrong. What set of experiments can you perform to locate the problem? Or, what experiments can you perform that don't require buying expensive gear?

My first suggestion is for you run a standalone GPS timing receiver like a M12+T or ublox-6. Put all your GPSDO away. Just use that OEM board, and your TIC, and your PRS-45A. Collect sawtooth data too. You should get down to 2 or 3 ns RMS.

/tvb

----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement" time-nuts@febo.com; "Magnus Danielson" magnus@rubidium.se
Sent: Thursday, April 27, 2017 9:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Magnus,
Try as I might, the weather and the local power company had other ideas about my long term capture. I'm running everything but the 5370 from a UPS. I guess I'm going to have to get batteries for my other UPS and run the 5370 from that. A one second power loss was all it took to stop the test.

Anyway, I did manage to get 376,238 points of data. The data is captured on a 5370A. The external clock input and the STOP channel are fed by the 10MHz from my PRS-45A. The START channel is fed by the 10MHz from one of my GPSDOs. The EXT channel is fed by the 1PPS from another of my GPSDO units. "EXT ARM" is enabled. So, essentially, at every 1PPS pulse, the phase difference between the two 10MHz feeds is captured.

I've attached a screenshot of the phase plot which can also be found here:http://evoria.net/AE6RV/Timelab/Screenshot.png
I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

So, back to my question: Where are the large ionospheric phase moves? This question has been causing me doubt since I started on this project. Or don't I still have enough data collected for this to happen?

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

  From: Magnus Danielson <magnus@rubidium.dyndns.org>

To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Tuesday, April 18, 2017 1:09 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Bob,

That is a good solution indeed. Good luck with that measurement run!

One of the fun stuff with Timelab is that you can walk by and check the
developments. I've found that very useful for long measurements (as in
hours and days).

I prepared a cesium for one vendor, and initially they did not care so
much, but then they saw more deviations between the receivers, so they
wanted to sort it out, but discovered that they could not cancel out the
common mode of GPS signals (and its shifts), so then firing up that
cesium was the right thing. I remember writing support emails while
waiting for the airplane in Madrid airport, happy that they was doing a
first run for the right measurement reason. :)

Cheers,
Magnus

On 04/18/2017 04:25 AM, Bob Stewart wrote:

Hi Tom,
Could you clarify the term RMS?  Is there some sort of calculation I need to do on the raw data to get the RMS value?
I'm going to begin by doing a fresh 8 hour test of GPSDO Unit 1 vs GPSDO Unit 2, with the same conditions as the tests involving the PRS.  IOW, the 5370 is being clocked by GPSDO Unit 2, and the trigger for the EXT input to the 5370 is from GPSDO Unit 2.  START is from GSPDO Unit 1 and STOP is from GPSDO Unit 2.  The ones I have on hand show about 4E-11 at 1s tau. 

After that, I'll run an 8 hour test of raw 1PPS from a LEA-6T against the PRS.  Since I don't really trust the PRS, I'll use GPSDO Unit 2 to clock the 5370.  If you really don't want the GSPDO to clock the 5370, I'll run the same test using the internal 10811 as the clock, to compare the difference.

I also ran into an old test of the 5370 where START (10MHz) was fed by a GPSDO and STOP (10MHz) was fed by the same GPSDO but with (I think) a 20 ft cable from START to STOP.  I'll probably want to rerun that, as well, but it was an almost straight line from 2E-11 at 1s tau in that dataset.  Still, that test will verify that the 5370 hasn't degraded. Bob

  From: Tom Van Baak <tvb@LeapSecond.com>

To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 12:02 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi BobS,

Thanks for sharing that. To follow-up on recent emails...

When you start to push the limits of your own test equipment it's good to employ tricks such as 3-corner hat in order to get a few dB better measurement. But there's a limit to that. And you probably don't want me to suggest that you run 3 GPSDO, 3 counters and 3 cesiums all simultaneously.

I've attached two plots showing 4+ days of your GPSDO data and 4+ days of a TBolt (factory defaults, untuned)

bob-gpsdo-3.gif -- Notice how much your GPSDO (blue trace) wanders all over the place.

bob-gpsdo-5.gif -- ADEV comparison. See below for comments.

When I look at the ADEV for your data set it's pretty clear that the limiting factor on the left (short tau) is the resolution of your 5370A counter. That's why the TBolt looks better. Don't worry about that.

In the middle of the ADEV plot it would appear that both your 5370A counter and your PRS-45A cesium reference are sufficiently good to properly measure your GPSDO. That's good. The TBolt looks bad in the middle but that's simply because it was using a way-too-short default time constant. With proper tuning the red trace would be mostly flat out to 1000 seconds.

http://leapsecond.com/pages/tbolt-8d/sigma1.gif from http://www.leapsecond.com/pages/tbolt-8d/
you'll note the sidereal "blip" is below 4e-14. Your ADEV doesn't get near that level.

I say this because a rule-of-thumb is that a good M12+T timing receiver should get you down to 3 ns per day RMS. And a good ublox-6T can get you to 3 or 2 ns per day RMS. Your data shows something more like 5 ns -- which for a GPSDO is worse than no GPSDO at all. So something's wrong. What set of experiments can you perform to locate the problem? Or, what experiments can you perform that don't require buying expensive gear?

My first suggestion is for you run a standalone GPS timing receiver like a M12+T or ublox-6. Put all your GPSDO away. Just use that OEM board, and your TIC, and your PRS-45A. Collect sawtooth data too. You should get down to 2 or 3 ns RMS.

/tvb

----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement" time-nuts@febo.com; "Magnus Danielson" magnus@rubidium.se
Sent: Thursday, April 27, 2017 9:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Magnus,
Try as I might, the weather and the local power company had other ideas about my long term capture. I'm running everything but the 5370 from a UPS. I guess I'm going to have to get batteries for my other UPS and run the 5370 from that. A one second power loss was all it took to stop the test.

Anyway, I did manage to get 376,238 points of data. The data is captured on a 5370A. The external clock input and the STOP channel are fed by the 10MHz from my PRS-45A. The START channel is fed by the 10MHz from one of my GPSDOs. The EXT channel is fed by the 1PPS from another of my GPSDO units. "EXT ARM" is enabled. So, essentially, at every 1PPS pulse, the phase difference between the two 10MHz feeds is captured.

I've attached a screenshot of the phase plot which can also be found here:http://evoria.net/AE6RV/Timelab/Screenshot.png
I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

So, back to my question: Where are the large ionospheric phase moves? This question has been causing me doubt since I started on this project. Or don't I still have enough data collected for this to happen?

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

      From: Magnus Danielson magnus@rubidium.dyndns.org
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Tuesday, April 18, 2017 1:09 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?
 
Hi Bob,

That is a good solution indeed. Good luck with that measurement run!

One of the fun stuff with Timelab is that you can walk by and check the
developments. I've found that very useful for long measurements (as in
hours and days).

I prepared a cesium for one vendor, and initially they did not care so
much, but then they saw more deviations between the receivers, so they
wanted to sort it out, but discovered that they could not cancel out the
common mode of GPS signals (and its shifts), so then firing up that
cesium was the right thing. I remember writing support emails while
waiting for the airplane in Madrid airport, happy that they was doing a
first run for the right measurement reason. :)

Cheers,
Magnus

On 04/18/2017 04:25 AM, Bob Stewart wrote:

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.

Hi

RMS:

https://en.wikipedia.org/wiki/Root_mean_square https://en.wikipedia.org/wiki/Root_mean_square

When we talk about things like “6 sigma” in QA, the sigma is the RMS.

Bob

Hi Bob,
What I should have asked is this: "how do I get that from a Timelab dataset?"   Bob

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: Tom Van Baak tvb@leapsecond.com
Sent: Wednesday, May 3, 2017 7:20 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
RMS:
https://en.wikipedia.org/wiki/Root_mean_square
When we talk about things like “6 sigma” in QA, the sigma is the RMS. 
Bob

On May 3, 2017, at 1:44 AM, Bob Stewart bob@evoria.net wrote:
Hi Tom,
Could you clarify the term RMS?  Is there some sort of calculation I need to do on the raw data to get the RMS value?
I'm going to begin by doing a fresh 8 hour test of GPSDO Unit 1 vs GPSDO Unit 2, with the same conditions as the tests involving the PRS.  IOW, the 5370 is being clocked by GPSDO Unit 2, and the trigger for the EXT input to the 5370 is from GPSDO Unit 2.  START is from GSPDO Unit 1 and STOP is from GPSDO Unit 2.  The ones I have on hand show about 4E-11 at 1s tau. 

After that, I'll run an 8 hour test of raw 1PPS from a LEA-6T against the PRS.  Since I don't really trust the PRS, I'll use GPSDO Unit 2 to clock the 5370.  If you really don't want the GSPDO to clock the 5370, I'll run the same test using the internal 10811 as the clock, to compare the difference.

I also ran into an old test of the 5370 where START (10MHz) was fed by a GPSDO and STOP (10MHz) was fed by the same GPSDO but with (I think) a 20 ft cable from START to STOP.  I'll probably want to rerun that, as well, but it was an almost straight line from 2E-11 at 1s tau in that dataset.  Still, that test will verify that the 5370 hasn't degraded. Bob

     From: Tom Van Baak tvb@LeapSecond.com
To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 12:02 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi BobS,

I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

Thanks for sharing that. To follow-up on recent emails...

When you start to push the limits of your own test equipment it's good to employ tricks such as 3-corner hat in order to get a few dB better measurement. But there's a limit to that. And you probably don't want me to suggest that you run 3 GPSDO, 3 counters and 3 cesiums all simultaneously.

I've attached two plots showing 4+ days of your GPSDO data and 4+ days of a TBolt (factory defaults, untuned)

bob-gpsdo-3.gif -- Notice how much your GPSDO (blue trace) wanders all over the place.

bob-gpsdo-5.gif -- ADEV comparison. See below for comments.

When I look at the ADEV for your data set it's pretty clear that the limiting factor on the left (short tau) is the resolution of your 5370A counter. That's why the TBolt looks better. Don't worry about that.

In the middle of the ADEV plot it would appear that both your 5370A counter and your PRS-45A cesium reference are sufficiently good to properly measure your GPSDO. That's good. The TBolt looks bad in the middle but that's simply because it was using a way-too-short default time constant. With proper tuning the red trace would be mostly flat out to 1000 seconds.

http://leapsecond.com/pages/tbolt-8d/sigma1.gif from http://www.leapsecond.com/pages/tbolt-8d/
you'll note the sidereal "blip" is below 4e-14. Your ADEV doesn't get near that level.

I say this because a rule-of-thumb is that a good M12+T timing receiver should get you down to 3 ns per day RMS. And a good ublox-6T can get you to 3 or 2 ns per day RMS. Your data shows something more like 5 ns -- which for a GPSDO is worse than no GPSDO at all. So something's wrong. What set of experiments can you perform to locate the problem? Or, what experiments can you perform that don't require buying expensive gear?

My first suggestion is for you run a standalone GPS timing receiver like a M12+T or ublox-6. Put all your GPSDO away. Just use that OEM board, and your TIC, and your PRS-45A. Collect sawtooth data too. You should get down to 2 or 3 ns RMS.

/tvb

----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement" time-nuts@febo.com; "Magnus Danielson" magnus@rubidium.se
Sent: Thursday, April 27, 2017 9:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Magnus,
Try as I might, the weather and the local power company had other ideas about my long term capture. I'm running everything but the 5370 from a UPS. I guess I'm going to have to get batteries for my other UPS and run the 5370 from that. A one second power loss was all it took to stop the test.

Anyway, I did manage to get 376,238 points of data. The data is captured on a 5370A. The external clock input and the STOP channel are fed by the 10MHz from my PRS-45A. The START channel is fed by the 10MHz from one of my GPSDOs. The EXT channel is fed by the 1PPS from another of my GPSDO units. "EXT ARM" is enabled. So, essentially, at every 1PPS pulse, the phase difference between the two 10MHz feeds is captured.

I've attached a screenshot of the phase plot which can also be found here:http://evoria.net/AE6RV/Timelab/Screenshot.png
I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

So, back to my question: Where are the large ionospheric phase moves? This question has been causing me doubt since I started on this project. Or don't I still have enough data collected for this to happen?

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

      From: Magnus Danielson magnus@rubidium.dyndns.org
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Tuesday, April 18, 2017 1:09 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?
 
Hi Bob,

That is a good solution indeed. Good luck with that measurement run!

One of the fun stuff with Timelab is that you can walk by and check the
developments. I've found that very useful for long measurements (as in
hours and days).

I prepared a cesium for one vendor, and initially they did not care so
much, but then they saw more deviations between the receivers, so they
wanted to sort it out, but discovered that they could not cancel out the
common mode of GPS signals (and its shifts), so then firing up that
cesium was the right thing. I remember writing support emails while
waiting for the airplane in Madrid airport, happy that they was doing a
first run for the right measurement reason. :)

Cheers,
Magnus

On 04/18/2017 04:25 AM, Bob Stewart wrote:

Hi Magnus,
Today I started a long run against my PRS-45A. Maybe this time I won't have a power outage. I'll see what it tells me in a few days.
Bob_______________________________________________

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

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.

Hi

Dump it to a csv file and do it in Excel would be one way. You want to be able
to control the process a bit so that would give you the ability to tweak the measurement.

Bob

Hi,

RMS is really a standard deviation estimation in classical context. This
type of deviation estimation is however not useful for the types of
noises we have, so that is why we needed a more powerful tool, and we
ended up with the Allan Deviation in its place.

Just as with RMS being the estimator formula for standard deviation,
there exist several estimator formulas for the Allan Deviation.

Cheers,
Magnus

On 05/03/2017 02:12 PM, Bob kb8tq wrote:

Hi Bob,
Won't I have a problem with the phase wrap unless I convert the data to a range of +50ns to -50ns?  Or am I still looking at this incorrectly?
 Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net
Cc: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 10:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
Dump it to a csv file and do it in Excel would be one way. You want to be ableto control the process a bit so that would give you the ability to tweak the measurement.
Bob

On May 3, 2017, at 11:01 AM, Bob Stewart bob@evoria.net wrote:
Hi Bob,
What I should have asked is this: "how do I get that from a Timelab dataset?"   Bob

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: Tom Van Baak tvb@leapsecond.com
Sent: Wednesday, May 3, 2017 7:20 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
RMS:
https://en.wikipedia.org/wiki/Root_mean_square
When we talk about things like “6 sigma” in QA, the sigma is the RMS. 
Bob

On May 3, 2017, at 1:44 AM, Bob Stewart bob@evoria.net wrote:
Hi Tom,
Could you clarify the term RMS?  Is there some sort of calculation I need to do on the raw data to get the RMS value?
I'm going to begin by doing a fresh 8 hour test of GPSDO Unit 1 vs GPSDO Unit 2, with the same conditions as the tests involving the PRS.  IOW, the 5370 is being clocked by GPSDO Unit 2, and the trigger for the EXT input to the 5370 is from GPSDO Unit 2.  START is from GSPDO Unit 1 and STOP is from GPSDO Unit 2.  The ones I have on hand show about 4E-11 at 1s tau. 

After that, I'll run an 8 hour test of raw 1PPS from a LEA-6T against the PRS.  Since I don't really trust the PRS, I'll use GPSDO Unit 2 to clock the 5370.  If you really don't want the GSPDO to clock the 5370, I'll run the same test using the internal 10811 as the clock, to compare the difference.

I also ran into an old test of the 5370 where START (10MHz) was fed by a GPSDO and STOP (10MHz) was fed by the same GPSDO but with (I think) a 20 ft cable from START to STOP.  I'll probably want to rerun that, as well, but it was an almost straight line from 2E-11 at 1s tau in that dataset.  Still, that test will verify that the 5370 hasn't degraded. Bob

     From: Tom Van Baak tvb@LeapSecond.com
To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 12:02 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi BobS,

I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

Thanks for sharing that. To follow-up on recent emails...

When you start to push the limits of your own test equipment it's good to employ tricks such as 3-corner hat in order to get a few dB better measurement. But there's a limit to that. And you probably don't want me to suggest that you run 3 GPSDO, 3 counters and 3 cesiums all simultaneously.

I've attached two plots showing 4+ days of your GPSDO data and 4+ days of a TBolt (factory defaults, untuned)

bob-gpsdo-3.gif -- Notice how much your GPSDO (blue trace) wanders all over the place.

bob-gpsdo-5.gif -- ADEV comparison. See below for comments.

When I look at the ADEV for your data set it's pretty clear that the limiting factor on the left (short tau) is the resolution of your 5370A counter. That's why the TBolt looks better. Don't worry about that.

In the middle of the ADEV plot it would appear that both your 5370A counter and your PRS-45A cesium reference are sufficiently good to properly measure your GPSDO. That's good. The TBolt looks bad in the middle but that's simply because it was using a way-too-short default time constant. With proper tuning the red trace would be mostly flat out to 1000 seconds.

http://leapsecond.com/pages/tbolt-8d/sigma1.gif from http://www.leapsecond.com/pages/tbolt-8d/
you'll note the sidereal "blip" is below 4e-14. Your ADEV doesn't get near that level.

I say this because a rule-of-thumb is that a good M12+T timing receiver should get you down to 3 ns per day RMS. And a good ublox-6T can get you to 3 or 2 ns per day RMS. Your data shows something more like 5 ns -- which for a GPSDO is worse than no GPSDO at all. So something's wrong. What set of experiments can you perform to locate the problem? Or, what experiments can you perform that don't require buying expensive gear?

My first suggestion is for you run a standalone GPS timing receiver like a M12+T or ublox-6. Put all your GPSDO away. Just use that OEM board, and your TIC, and your PRS-45A. Collect sawtooth data too. You should get down to 2 or 3 ns RMS.

/tvb

----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement" time-nuts@febo.com; "Magnus Danielson" magnus@rubidium.se
Sent: Thursday, April 27, 2017 9:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Magnus,
Try as I might, the weather and the local power company had other ideas about my long term capture. I'm running everything but the 5370 from a UPS. I guess I'm going to have to get batteries for my other UPS and run the 5370 from that. A one second power loss was all it took to stop the test.

Anyway, I did manage to get 376,238 points of data. The data is captured on a 5370A. The external clock input and the STOP channel are fed by the 10MHz from my PRS-45A. The START channel is fed by the 10MHz from one of my GPSDOs. The EXT channel is fed by the 1PPS from another of my GPSDO units. "EXT ARM" is enabled. So, essentially, at every 1PPS pulse, the phase difference between the two 10MHz feeds is captured.

I've attached a screenshot of the phase plot which can also be found here:http://evoria.net/AE6RV/Timelab/Screenshot.png
I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

So, back to my question: Where are the large ionospheric phase moves? This question has been causing me doubt since I started on this project. Or don't I still have enough data collected for this to happen?

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

      From: Magnus Danielson magnus@rubidium.dyndns.org
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Tuesday, April 18, 2017 1:09 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?
 
Hi Bob,

That is a good solution indeed. Good luck with that measurement run!

One of the fun stuff with Timelab is that you can walk by and check the
developments. I've found that very useful for long measurements (as in
hours and days).

I prepared a cesium for one vendor, and initially they did not care so
much, but then they saw more deviations between the receivers, so they
wanted to sort it out, but discovered that they could not cancel out the
common mode of GPS signals (and its shifts), so then firing up that
cesium was the right thing. I remember writing support emails while
waiting for the airplane in Madrid airport, happy that they was doing a
first run for the right measurement reason. :)

Cheers,
Magnus

On 04/18/2017 04:25 AM, Bob Stewart wrote:

Hi Magnus,
Today I started a long run against my PRS-45A. Maybe this time I won't have a power outage. I'll see what it tells me in a few days.
Bob_______________________________________________

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

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.

Hi

If you convert the data to +/- 50 ns, you will have a problem at the edges. Much easier to deal with
unwrapped data. The same CSV approach works for taking the pps data from a module, logging
the sawtooth info and combining them. There’s no practical way to get the “real” RMS of the module
without doing that.

Bob

Hi Bob,
OK, I think I see my error.  I need to create a file of the successive phase differences in the timelab capture, right?  I haven't done this before, so...

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net
Cc: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 11:11 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
If you convert the data to +/- 50 ns, you will have a problem at the edges. Much easier to deal with unwrapped data. The same CSV approach works for taking the pps data from a module, logging the sawtooth info and combining them. There’s no practical way to get the “real” RMS of the module without doing that. 
Bob

On May 3, 2017, at 11:56 AM, Bob Stewart bob@evoria.net wrote:
Hi Bob,
Won't I have a problem with the phase wrap unless I convert the data to a range of +50ns to -50ns?  Or am I still looking at this incorrectly?
 Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net
Cc: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 10:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
Dump it to a csv file and do it in Excel would be one way. You want to be ableto control the process a bit so that would give you the ability to tweak the measurement.
Bob

On May 3, 2017, at 11:01 AM, Bob Stewart bob@evoria.net wrote:
Hi Bob,
What I should have asked is this: "how do I get that from a Timelab dataset?"   Bob

  From: Bob kb8tq <kb8tq@n1k.org>

To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: Tom Van Baak tvb@leapsecond.com
Sent: Wednesday, May 3, 2017 7:20 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi
RMS:
https://en.wikipedia.org/wiki/Root_mean_square
When we talk about things like “6 sigma” in QA, the sigma is the RMS. 
Bob

On May 3, 2017, at 1:44 AM, Bob Stewart bob@evoria.net wrote:
Hi Tom,
Could you clarify the term RMS?  Is there some sort of calculation I need to do on the raw data to get the RMS value?
I'm going to begin by doing a fresh 8 hour test of GPSDO Unit 1 vs GPSDO Unit 2, with the same conditions as the tests involving the PRS.  IOW, the 5370 is being clocked by GPSDO Unit 2, and the trigger for the EXT input to the 5370 is from GPSDO Unit 2.  START is from GSPDO Unit 1 and STOP is from GPSDO Unit 2.  The ones I have on hand show about 4E-11 at 1s tau. 

After that, I'll run an 8 hour test of raw 1PPS from a LEA-6T against the PRS.  Since I don't really trust the PRS, I'll use GPSDO Unit 2 to clock the 5370.  If you really don't want the GSPDO to clock the 5370, I'll run the same test using the internal 10811 as the clock, to compare the difference.

I also ran into an old test of the 5370 where START (10MHz) was fed by a GPSDO and STOP (10MHz) was fed by the same GPSDO but with (I think) a 20 ft cable from START to STOP.  I'll probably want to rerun that, as well, but it was an almost straight line from 2E-11 at 1s tau in that dataset.  Still, that test will verify that the 5370 hasn't degraded. Bob

     From: Tom Van Baak tvb@LeapSecond.com
To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 12:02 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi BobS,

I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

Thanks for sharing that. To follow-up on recent emails...

When you start to push the limits of your own test equipment it's good to employ tricks such as 3-corner hat in order to get a few dB better measurement. But there's a limit to that. And you probably don't want me to suggest that you run 3 GPSDO, 3 counters and 3 cesiums all simultaneously.

I've attached two plots showing 4+ days of your GPSDO data and 4+ days of a TBolt (factory defaults, untuned)

bob-gpsdo-3.gif -- Notice how much your GPSDO (blue trace) wanders all over the place.

bob-gpsdo-5.gif -- ADEV comparison. See below for comments.

When I look at the ADEV for your data set it's pretty clear that the limiting factor on the left (short tau) is the resolution of your 5370A counter. That's why the TBolt looks better. Don't worry about that.

In the middle of the ADEV plot it would appear that both your 5370A counter and your PRS-45A cesium reference are sufficiently good to properly measure your GPSDO. That's good. The TBolt looks bad in the middle but that's simply because it was using a way-too-short default time constant. With proper tuning the red trace would be mostly flat out to 1000 seconds.

http://leapsecond.com/pages/tbolt-8d/sigma1.gif from http://www.leapsecond.com/pages/tbolt-8d/
you'll note the sidereal "blip" is below 4e-14. Your ADEV doesn't get near that level.

I say this because a rule-of-thumb is that a good M12+T timing receiver should get you down to 3 ns per day RMS. And a good ublox-6T can get you to 3 or 2 ns per day RMS. Your data shows something more like 5 ns -- which for a GPSDO is worse than no GPSDO at all. So something's wrong. What set of experiments can you perform to locate the problem? Or, what experiments can you perform that don't require buying expensive gear?

My first suggestion is for you run a standalone GPS timing receiver like a M12+T or ublox-6. Put all your GPSDO away. Just use that OEM board, and your TIC, and your PRS-45A. Collect sawtooth data too. You should get down to 2 or 3 ns RMS.

/tvb

----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement" time-nuts@febo.com; "Magnus Danielson" magnus@rubidium.se
Sent: Thursday, April 27, 2017 9:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Magnus,
Try as I might, the weather and the local power company had other ideas about my long term capture. I'm running everything but the 5370 from a UPS. I guess I'm going to have to get batteries for my other UPS and run the 5370 from that. A one second power loss was all it took to stop the test.

Anyway, I did manage to get 376,238 points of data. The data is captured on a 5370A. The external clock input and the STOP channel are fed by the 10MHz from my PRS-45A. The START channel is fed by the 10MHz from one of my GPSDOs. The EXT channel is fed by the 1PPS from another of my GPSDO units. "EXT ARM" is enabled. So, essentially, at every 1PPS pulse, the phase difference between the two 10MHz feeds is captured.

I've attached a screenshot of the phase plot which can also be found here:http://evoria.net/AE6RV/Timelab/Screenshot.png
I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

So, back to my question: Where are the large ionospheric phase moves? This question has been causing me doubt since I started on this project. Or don't I still have enough data collected for this to happen?

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

      From: Magnus Danielson magnus@rubidium.dyndns.org
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Tuesday, April 18, 2017 1:09 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?
 
Hi Bob,

That is a good solution indeed. Good luck with that measurement run!

One of the fun stuff with Timelab is that you can walk by and check the
developments. I've found that very useful for long measurements (as in
hours and days).

I prepared a cesium for one vendor, and initially they did not care so
much, but then they saw more deviations between the receivers, so they
wanted to sort it out, but discovered that they could not cancel out the
common mode of GPS signals (and its shifts), so then firing up that
cesium was the right thing. I remember writing support emails while
waiting for the airplane in Madrid airport, happy that they was doing a
first run for the right measurement reason. :)

Cheers,
Magnus

On 04/18/2017 04:25 AM, Bob Stewart wrote:

Hi Magnus,
Today I started a long run against my PRS-45A. Maybe this time I won't have a power outage. I'll see what it tells me in a few days.
Bob_______________________________________________

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

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.

Hi Tom et al,
It looks like I've got a problem with my 5370 when collecting 10MHz data using an external arming signal.  I did some tests today, sending 10MHz through a ~10ft delay line (about 8.26ns) between the START and STOP inputs.  It's a mess, with some interval deltas being as high as 13 ns.  I tested with both an external clock and the internal 10811 clock with essentially the same result.
I'm running a test right now using a 1PPS signal and the delay line with no external arming signal, and, so far, there's no problem.
So, back to the drawing board.

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

  From: Tom Van Baak <tvb@LeapSecond.com>

To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Wednesday, May 3, 2017 12:02 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi BobS,

Thanks for sharing that. To follow-up on recent emails...

When you start to push the limits of your own test equipment it's good to employ tricks such as 3-corner hat in order to get a few dB better measurement. But there's a limit to that. And you probably don't want me to suggest that you run 3 GPSDO, 3 counters and 3 cesiums all simultaneously.

I've attached two plots showing 4+ days of your GPSDO data and 4+ days of a TBolt (factory defaults, untuned)

bob-gpsdo-3.gif -- Notice how much your GPSDO (blue trace) wanders all over the place.

bob-gpsdo-5.gif -- ADEV comparison. See below for comments.

When I look at the ADEV for your data set it's pretty clear that the limiting factor on the left (short tau) is the resolution of your 5370A counter. That's why the TBolt looks better. Don't worry about that.

In the middle of the ADEV plot it would appear that both your 5370A counter and your PRS-45A cesium reference are sufficiently good to properly measure your GPSDO. That's good. The TBolt looks bad in the middle but that's simply because it was using a way-too-short default time constant. With proper tuning the red trace would be mostly flat out to 1000 seconds.

http://leapsecond.com/pages/tbolt-8d/sigma1.gif from http://www.leapsecond.com/pages/tbolt-8d/
you'll note the sidereal "blip" is below 4e-14. Your ADEV doesn't get near that level.

I say this because a rule-of-thumb is that a good M12+T timing receiver should get you down to 3 ns per day RMS. And a good ublox-6T can get you to 3 or 2 ns per day RMS. Your data shows something more like 5 ns -- which for a GPSDO is worse than no GPSDO at all. So something's wrong. What set of experiments can you perform to locate the problem? Or, what experiments can you perform that don't require buying expensive gear?

My first suggestion is for you run a standalone GPS timing receiver like a M12+T or ublox-6. Put all your GPSDO away. Just use that OEM board, and your TIC, and your PRS-45A. Collect sawtooth data too. You should get down to 2 or 3 ns RMS.

/tvb

----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement" time-nuts@febo.com; "Magnus Danielson" magnus@rubidium.se
Sent: Thursday, April 27, 2017 9:48 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?

Hi Magnus,
Try as I might, the weather and the local power company had other ideas about my long term capture. I'm running everything but the 5370 from a UPS. I guess I'm going to have to get batteries for my other UPS and run the 5370 from that. A one second power loss was all it took to stop the test.

Anyway, I did manage to get 376,238 points of data. The data is captured on a 5370A. The external clock input and the STOP channel are fed by the 10MHz from my PRS-45A. The START channel is fed by the 10MHz from one of my GPSDOs. The EXT channel is fed by the 1PPS from another of my GPSDO units. "EXT ARM" is enabled. So, essentially, at every 1PPS pulse, the phase difference between the two 10MHz feeds is captured.

I've attached a screenshot of the phase plot which can also be found here:http://evoria.net/AE6RV/Timelab/Screenshot.png
I've also made the timelab file (compressed by 7z) available here:
http://evoria.net/AE6RV/Timelab/GFSvsCS.4.22.17.7z

So, back to my question: Where are the large ionospheric phase moves? This question has been causing me doubt since I started on this project. Or don't I still have enough data collected for this to happen?

Bob

AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

      From: Magnus Danielson magnus@rubidium.dyndns.org
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Tuesday, April 18, 2017 1:09 AM
Subject: Re: [time-nuts] Three-cornered hat on timelab?
 
Hi Bob,

That is a good solution indeed. Good luck with that measurement run!

One of the fun stuff with Timelab is that you can walk by and check the
developments. I've found that very useful for long measurements (as in
hours and days).

I prepared a cesium for one vendor, and initially they did not care so
much, but then they saw more deviations between the receivers, so they
wanted to sort it out, but discovered that they could not cancel out the
common mode of GPS signals (and its shifts), so then firing up that
cesium was the right thing. I remember writing support emails while
waiting for the airplane in Madrid airport, happy that they was doing a
first run for the right measurement reason. :)

Cheers,
Magnus

On 04/18/2017 04:25 AM, Bob Stewart wrote:

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.