Empathy List Archives

DR

Doug Ronald

Fri, Mar 3, 2023 7:51 PM

I'm back for a new (to me) question:
I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal oscillator and the other with a rubidium
oscillator. Both are receiving from the same GPS antenna, and both receivers are tracking 5 or 6 satellites all the
time. I repaired the rubidium clock itself where the oven was too hot to maintain internal lock, and has an external VCO
line to pull the unit to the GPS 1 PPS. That line has a correction voltage that changes, then stabilizes to about 2.8
VDC.
With an oscilloscope, I monitor the phase of the 10 MHz from one unit to the other, and much to my surprise, the phase
drifts, always in the same direction, but not linearly. It takes about 40 minutes to complete a 360 degree cycle. I
expected there would maybe be some tiny jitter in the phase, but not a continuous error.
My fear is that one or both of the 10 MHz references is not being corrected to the GPS signal, but I thought I would ask
here to see if this drift might be expected. Probably others have done this experiment, so I think I'll get a good
answer.
Thanks,
Doug Ronald, W6DSR

I'm back for a new (to me) question: I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal oscillator and the other with a rubidium oscillator. Both are receiving from the same GPS antenna, and both receivers are tracking 5 or 6 satellites all the time. I repaired the rubidium clock itself where the oven was too hot to maintain internal lock, and has an external VCO line to pull the unit to the GPS 1 PPS. That line has a correction voltage that changes, then stabilizes to about 2.8 VDC. With an oscilloscope, I monitor the phase of the 10 MHz from one unit to the other, and much to my surprise, the phase drifts, always in the same direction, but not linearly. It takes about 40 minutes to complete a 360 degree cycle. I expected there would maybe be some tiny jitter in the phase, but not a continuous error. My fear is that one or both of the 10 MHz references is not being corrected to the GPS signal, but I thought I would ask here to see if this drift might be expected. Probably others have done this experiment, so I think I'll get a good answer. Thanks, Doug Ronald, W6DSR

BC

Bob Camp

Sat, Mar 4, 2023 2:40 AM

Hi

If you see multiple cycle slips and they are in the same direction, then yes, at least one unit
is unlocked.

~40 minutes is about 2,400 seconds. One cycle at 10 MHz should be 100 ns. More or less
you are slipping 1/24 ns per second. It’s not a lot.

A way to double check things would be to take a look at the PPS outputs. Even a pretty
basic setup should be able to resolve < 100 ns. An old HP 5336 era counter off eBay might
be a sub $100 sort of option. Any of a number of MCU based boards could do the job
with a bit of code. A digital scope would be anoter option.

Bob

On Mar 3, 2023, at 2:51 PM, Doug Ronald via time-nuts time-nuts@lists.febo.com wrote:

I'm back for a new (to me) question:
I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal oscillator and the other with a rubidium
oscillator. Both are receiving from the same GPS antenna, and both receivers are tracking 5 or 6 satellites all the
time. I repaired the rubidium clock itself where the oven was too hot to maintain internal lock, and has an external VCO
line to pull the unit to the GPS 1 PPS. That line has a correction voltage that changes, then stabilizes to about 2.8
VDC.
With an oscilloscope, I monitor the phase of the 10 MHz from one unit to the other, and much to my surprise, the phase
drifts, always in the same direction, but not linearly. It takes about 40 minutes to complete a 360 degree cycle. I
expected there would maybe be some tiny jitter in the phase, but not a continuous error.
My fear is that one or both of the 10 MHz references is not being corrected to the GPS signal, but I thought I would ask
here to see if this drift might be expected. Probably others have done this experiment, so I think I'll get a good
answer.
Thanks,
Doug Ronald, W6DSR

time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com

Hi If you see multiple cycle slips and they are in the same direction, then yes, at least one unit is unlocked. ~40 minutes is about 2,400 seconds. One cycle at 10 MHz should be 100 ns. More or less you are slipping 1/24 ns per second. It’s not a lot. A way to double check things would be to take a look at the PPS outputs. Even a pretty basic setup should be able to resolve < 100 ns. An old HP 5336 era counter off eBay might be a sub $100 sort of option. Any of a number of MCU based boards could do the job with a bit of code. A digital scope would be anoter option. Bob > On Mar 3, 2023, at 2:51 PM, Doug Ronald via time-nuts <time-nuts@lists.febo.com> wrote: > > I'm back for a new (to me) question: > I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal oscillator and the other with a rubidium > oscillator. Both are receiving from the same GPS antenna, and both receivers are tracking 5 or 6 satellites all the > time. I repaired the rubidium clock itself where the oven was too hot to maintain internal lock, and has an external VCO > line to pull the unit to the GPS 1 PPS. That line has a correction voltage that changes, then stabilizes to about 2.8 > VDC. > With an oscilloscope, I monitor the phase of the 10 MHz from one unit to the other, and much to my surprise, the phase > drifts, always in the same direction, but not linearly. It takes about 40 minutes to complete a 360 degree cycle. I > expected there would maybe be some tiny jitter in the phase, but not a continuous error. > My fear is that one or both of the 10 MHz references is not being corrected to the GPS signal, but I thought I would ask > here to see if this drift might be expected. Probably others have done this experiment, so I think I'll get a good > answer. > Thanks, > Doug Ronald, W6DSR > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com

K

Kitski

Sat, Mar 4, 2023 3:31 AM

Hi Doug,

Simple answer is, they should remain 'in phase'. That said, there may be a
few degrees of jitter/wander/whatever the locking mechanism does over the
medium term, but all should settle back to nominal phase 'coherence'. If I
understand you correctly, one/both of your DuT is not locked to 'mother'.
As usual, you'll need a third GPS to confirm ! Do your TrueTimes have an
interrogation port to see what the internals are doing ?

Kit
VK1LL

Hi Doug, Simple answer is, they should remain 'in phase'. That said, there may be a few degrees of jitter/wander/whatever the locking mechanism does over the medium term, but all should settle back to nominal phase 'coherence'. If I understand you correctly, one/both of your DuT is not locked to 'mother'. As usual, you'll need a third GPS to confirm ! Do your TrueTimes have an interrogation port to see what the internals are doing ? Kit VK1LL

DW

Dana Whitlow

Sat, Mar 4, 2023 10:21 AM

Doug,

Have you tried the corresponding experiment of looking for time drift
between the
two 1 PPS signals' leading edges? If so, what was the result?

Dana

On Fri, Mar 3, 2023 at 7:57 PM Doug Ronald via time-nuts <
time-nuts@lists.febo.com> wrote:

I'm back for a new (to me) question:
I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal
oscillator and the other with a rubidium
oscillator. Both are receiving from the same GPS antenna, and both
receivers are tracking 5 or 6 satellites all the
time. I repaired the rubidium clock itself where the oven was too hot to
maintain internal lock, and has an external VCO
line to pull the unit to the GPS 1 PPS. That line has a correction voltage
that changes, then stabilizes to about 2.8
VDC.
With an oscilloscope, I monitor the phase of the 10 MHz from one unit to
the other, and much to my surprise, the phase
drifts, always in the same direction, but not linearly. It takes about 40
minutes to complete a 360 degree cycle. I
expected there would maybe be some tiny jitter in the phase, but not a
continuous error.
My fear is that one or both of the 10 MHz references is not being
corrected to the GPS signal, but I thought I would ask
here to see if this drift might be expected. Probably others have done
this experiment, so I think I'll get a good
answer.
Thanks,
Doug Ronald, W6DSR

time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com

Doug, Have you tried the corresponding experiment of looking for time drift between the two 1 PPS signals' leading edges? If so, what was the result? Dana On Fri, Mar 3, 2023 at 7:57 PM Doug Ronald via time-nuts < time-nuts@lists.febo.com> wrote: > I'm back for a new (to me) question: > I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal > oscillator and the other with a rubidium > oscillator. Both are receiving from the same GPS antenna, and both > receivers are tracking 5 or 6 satellites all the > time. I repaired the rubidium clock itself where the oven was too hot to > maintain internal lock, and has an external VCO > line to pull the unit to the GPS 1 PPS. That line has a correction voltage > that changes, then stabilizes to about 2.8 > VDC. > With an oscilloscope, I monitor the phase of the 10 MHz from one unit to > the other, and much to my surprise, the phase > drifts, always in the same direction, but not linearly. It takes about 40 > minutes to complete a 360 degree cycle. I > expected there would maybe be some tiny jitter in the phase, but not a > continuous error. > My fear is that one or both of the 10 MHz references is not being > corrected to the GPS signal, but I thought I would ask > here to see if this drift might be expected. Probably others have done > this experiment, so I think I'll get a good > answer. > Thanks, > Doug Ronald, W6DSR > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com >

DR

Doug Ronald

Sat, Mar 4, 2023 7:29 PM

Thanks all for the responses.

I connected one XL-DC to the start trigger, and the other XL-DC to the stop trigger of an HP-5370B Time Interval Counter with the counter's clock on its own internal ovenized crystal oscillator. Interestingly, the two 1 PPS signals drift fairly constantly by about 300 ps/second. So, what I don't know is the source of the 1 PPS signals from the XL-DCs. Is that signal counted down from their internal references, or are they raw 1 PPS from the GPS receivers? I suspect they are simply counted down from the references, since they remain when I disconnect the GPS antenna, so I'm really not gaining any more insight than from measuring the 10 MHz oscillators - right?

I have a Lucent RFTG-U pair which I guess I'll have to get powered up for a 3rd reference to see which XL-DC is unlocked, but I strongly suspect the rubidium unit which I hate to have to attack again with such a complicated and difficult to work physical arrangement, especially in the physics box. I don't regularly use the RFTGs because they draw a lot of power, and radiate boo-coo RFI.

Thanks again for the ideas,
-Doug, W6DSR

-----Original Message-----
From: Dana Whitlow via time-nuts [mailto:time-nuts@lists.febo.com]
Sent: Saturday, March 4, 2023 02:22
To: Discussion of precise time and frequency measurement
Cc: Dana Whitlow
Subject: [time-nuts] Re: Two GPSDO 10 MHz phase drift

Doug,

Have you tried the corresponding experiment of looking for time drift
between the
two 1 PPS signals' leading edges? If so, what was the result?

Dana

On Fri, Mar 3, 2023 at 7:57 PM Doug Ronald via time-nuts <
time-nuts@lists.febo.com> wrote:

I'm back for a new (to me) question:
I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal
oscillator and the other with a rubidium
oscillator. Both are receiving from the same GPS antenna, and both
receivers are tracking 5 or 6 satellites all the
time. I repaired the rubidium clock itself where the oven was too hot to
maintain internal lock, and has an external VCO
line to pull the unit to the GPS 1 PPS. That line has a correction voltage
that changes, then stabilizes to about 2.8
VDC.
With an oscilloscope, I monitor the phase of the 10 MHz from one unit to
the other, and much to my surprise, the phase
drifts, always in the same direction, but not linearly. It takes about 40
minutes to complete a 360 degree cycle. I
expected there would maybe be some tiny jitter in the phase, but not a
continuous error.
My fear is that one or both of the 10 MHz references is not being
corrected to the GPS signal, but I thought I would ask
here to see if this drift might be expected. Probably others have done
this experiment, so I think I'll get a good
answer.
Thanks,
Doug Ronald, W6DSR

time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com

Thanks all for the responses. I connected one XL-DC to the start trigger, and the other XL-DC to the stop trigger of an HP-5370B Time Interval Counter with the counter's clock on its own internal ovenized crystal oscillator. Interestingly, the two 1 PPS signals drift fairly constantly by about 300 ps/second. So, what I don't know is the source of the 1 PPS signals from the XL-DCs. Is that signal counted down from their internal references, or are they raw 1 PPS from the GPS receivers? I suspect they are simply counted down from the references, since they remain when I disconnect the GPS antenna, so I'm really not gaining any more insight than from measuring the 10 MHz oscillators - right? I have a Lucent RFTG-U pair which I guess I'll have to get powered up for a 3rd reference to see which XL-DC is unlocked, but I strongly suspect the rubidium unit which I hate to have to attack again with such a complicated and difficult to work physical arrangement, especially in the physics box. I don't regularly use the RFTGs because they draw a lot of power, and radiate boo-coo RFI. Thanks again for the ideas, -Doug, W6DSR -----Original Message----- From: Dana Whitlow via time-nuts [mailto:time-nuts@lists.febo.com] Sent: Saturday, March 4, 2023 02:22 To: Discussion of precise time and frequency measurement Cc: Dana Whitlow Subject: [time-nuts] Re: Two GPSDO 10 MHz phase drift Doug, Have you tried the corresponding experiment of looking for time drift between the two 1 PPS signals' leading edges? If so, what was the result? Dana On Fri, Mar 3, 2023 at 7:57 PM Doug Ronald via time-nuts < time-nuts@lists.febo.com> wrote: > I'm back for a new (to me) question: > I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal > oscillator and the other with a rubidium > oscillator. Both are receiving from the same GPS antenna, and both > receivers are tracking 5 or 6 satellites all the > time. I repaired the rubidium clock itself where the oven was too hot to > maintain internal lock, and has an external VCO > line to pull the unit to the GPS 1 PPS. That line has a correction voltage > that changes, then stabilizes to about 2.8 > VDC. > With an oscilloscope, I monitor the phase of the 10 MHz from one unit to > the other, and much to my surprise, the phase > drifts, always in the same direction, but not linearly. It takes about 40 > minutes to complete a 360 degree cycle. I > expected there would maybe be some tiny jitter in the phase, but not a > continuous error. > My fear is that one or both of the 10 MHz references is not being > corrected to the GPS signal, but I thought I would ask > here to see if this drift might be expected. Probably others have done > this experiment, so I think I'll get a good > answer. > Thanks, > Doug Ronald, W6DSR > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com > _______________________________________________ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com

DR

Doug Ronald

Sun, Mar 5, 2023 1:43 AM

Hi all,

It's definitely the rubidium that isn't locking to GPS. I went back to 10 MHz with a Lissajous pattern on the o-scope, and with the GPS antenna pulled from the quartz XL-DC, that quartz unit drifted off so badly I wouldn't trust it to predict my bathroom schedule. I then let the quartz unit relieve a sigh of relief by connecting it back to its GPS teet, and pulled the rubidium unit off GPS. The 10 MHz went through a 360 degree shift in 3663 seconds. Putting it back on the GPS antenna, the drift didn't change, showing the rubidium is pretty good, but I would like it to be coherent with a GPS reference, so I guess I have my work is cut out for me in that Efratom box.

I kind 'a knew both units should be coherent, but wanted it confirmed, so thanks all for doing that...
-Doug Ronald, W6DSR
P.S. Why does it take so long to get a message posted?

-----Original Message-----
From: Doug Ronald [mailto:doug@dougronald.com]
Sent: Saturday, March 4, 2023 11:29
To: 'Discussion of precise time and frequency measurement'
Subject: RE: [time-nuts] Re: Two GPSDO 10 MHz phase drift

Thanks all for the responses.

I connected one XL-DC to the start trigger, and the other XL-DC to the stop trigger of an HP-5370B Time Interval Counter with the counter's clock on its own internal ovenized crystal oscillator. Interestingly, the two 1 PPS signals drift fairly constantly by about 300 ps/second. So, what I don't know is the source of the 1 PPS signals from the XL-DCs. Is that signal counted down from their internal references, or are they raw 1 PPS from the GPS receivers? I suspect they are simply counted down from the references, since they remain when I disconnect the GPS antenna, so I'm really not gaining any more insight than from measuring the 10 MHz oscillators - right?

I have a Lucent RFTG-U pair which I guess I'll have to get powered up for a 3rd reference to see which XL-DC is unlocked, but I strongly suspect the rubidium unit which I hate to have to attack again with such a complicated and difficult to work physical arrangement, especially in the physics box. I don't regularly use the RFTGs because they draw a lot of power, and radiate boo-coo RFI.

Thanks again for the ideas,
-Doug, W6DSR

-----Original Message-----
From: Dana Whitlow via time-nuts [mailto:time-nuts@lists.febo.com]
Sent: Saturday, March 4, 2023 02:22
To: Discussion of precise time and frequency measurement
Cc: Dana Whitlow
Subject: [time-nuts] Re: Two GPSDO 10 MHz phase drift

Doug,

Have you tried the corresponding experiment of looking for time drift
between the
two 1 PPS signals' leading edges? If so, what was the result?

Dana

On Fri, Mar 3, 2023 at 7:57 PM Doug Ronald via time-nuts <
time-nuts@lists.febo.com> wrote:

I'm back for a new (to me) question:
I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal
oscillator and the other with a rubidium
oscillator. Both are receiving from the same GPS antenna, and both
receivers are tracking 5 or 6 satellites all the
time. I repaired the rubidium clock itself where the oven was too hot to
maintain internal lock, and has an external VCO
line to pull the unit to the GPS 1 PPS. That line has a correction voltage
that changes, then stabilizes to about 2.8
VDC.
With an oscilloscope, I monitor the phase of the 10 MHz from one unit to
the other, and much to my surprise, the phase
drifts, always in the same direction, but not linearly. It takes about 40
minutes to complete a 360 degree cycle. I
expected there would maybe be some tiny jitter in the phase, but not a
continuous error.
My fear is that one or both of the 10 MHz references is not being
corrected to the GPS signal, but I thought I would ask
here to see if this drift might be expected. Probably others have done
this experiment, so I think I'll get a good
answer.
Thanks,
Doug Ronald, W6DSR

time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com

Hi all, It's definitely the rubidium that isn't locking to GPS. I went back to 10 MHz with a Lissajous pattern on the o-scope, and with the GPS antenna pulled from the quartz XL-DC, that quartz unit drifted off so badly I wouldn't trust it to predict my bathroom schedule. I then let the quartz unit relieve a sigh of relief by connecting it back to its GPS teet, and pulled the rubidium unit off GPS. The 10 MHz went through a 360 degree shift in 3663 seconds. Putting it back on the GPS antenna, the drift didn't change, showing the rubidium is pretty good, but I would like it to be coherent with a GPS reference, so I guess I have my work is cut out for me in that Efratom box. I kind 'a knew both units should be coherent, but wanted it confirmed, so thanks all for doing that... -Doug Ronald, W6DSR P.S. Why does it take so long to get a message posted? -----Original Message----- From: Doug Ronald [mailto:doug@dougronald.com] Sent: Saturday, March 4, 2023 11:29 To: 'Discussion of precise time and frequency measurement' Subject: RE: [time-nuts] Re: Two GPSDO 10 MHz phase drift Thanks all for the responses. I connected one XL-DC to the start trigger, and the other XL-DC to the stop trigger of an HP-5370B Time Interval Counter with the counter's clock on its own internal ovenized crystal oscillator. Interestingly, the two 1 PPS signals drift fairly constantly by about 300 ps/second. So, what I don't know is the source of the 1 PPS signals from the XL-DCs. Is that signal counted down from their internal references, or are they raw 1 PPS from the GPS receivers? I suspect they are simply counted down from the references, since they remain when I disconnect the GPS antenna, so I'm really not gaining any more insight than from measuring the 10 MHz oscillators - right? I have a Lucent RFTG-U pair which I guess I'll have to get powered up for a 3rd reference to see which XL-DC is unlocked, but I strongly suspect the rubidium unit which I hate to have to attack again with such a complicated and difficult to work physical arrangement, especially in the physics box. I don't regularly use the RFTGs because they draw a lot of power, and radiate boo-coo RFI. Thanks again for the ideas, -Doug, W6DSR -----Original Message----- From: Dana Whitlow via time-nuts [mailto:time-nuts@lists.febo.com] Sent: Saturday, March 4, 2023 02:22 To: Discussion of precise time and frequency measurement Cc: Dana Whitlow Subject: [time-nuts] Re: Two GPSDO 10 MHz phase drift Doug, Have you tried the corresponding experiment of looking for time drift between the two 1 PPS signals' leading edges? If so, what was the result? Dana On Fri, Mar 3, 2023 at 7:57 PM Doug Ronald via time-nuts < time-nuts@lists.febo.com> wrote: > I'm back for a new (to me) question: > I have two TrueTime XL-DC receivers, one with an ovenized 10 MHz crystal > oscillator and the other with a rubidium > oscillator. Both are receiving from the same GPS antenna, and both > receivers are tracking 5 or 6 satellites all the > time. I repaired the rubidium clock itself where the oven was too hot to > maintain internal lock, and has an external VCO > line to pull the unit to the GPS 1 PPS. That line has a correction voltage > that changes, then stabilizes to about 2.8 > VDC. > With an oscilloscope, I monitor the phase of the 10 MHz from one unit to > the other, and much to my surprise, the phase > drifts, always in the same direction, but not linearly. It takes about 40 > minutes to complete a 360 degree cycle. I > expected there would maybe be some tiny jitter in the phase, but not a > continuous error. > My fear is that one or both of the 10 MHz references is not being > corrected to the GPS signal, but I thought I would ask > here to see if this drift might be expected. Probably others have done > this experiment, so I think I'll get a good > answer. > Thanks, > Doug Ronald, W6DSR > > _______________________________________________ > time-nuts mailing list -- time-nuts@lists.febo.com > To unsubscribe send an email to time-nuts-leave@lists.febo.com > _______________________________________________ time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com