New 5370A
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
http://evoria.net/AE6RV/5370A/Test1.png
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
Bob
Hi Bob,
On 05/14/2015 05:11 AM, Bob Stewart wrote:
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
The wiggle indicates that you have some form of disturbance on top of
your signal. It could be hum, it could be the 5 MHz PWM noise that the
5370A/B clock buffer board produces. Anyway, that wiggle is typical of a
noise-source and ADEV isn't always the best tool to find out what it is.
A FFT would potentially be better for analyzing that issue.
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
Yes, it's a good tool and it's sad that it doesn't play well with Wine
in Linux. I think John pointed out that Wine did not treat a
line-drawing call correctly. Anyway, regardless of who is at fault, it's
sad they don't play well together.
Cheers,
Magnus
Bob Stewart wrote:
... And kudos to John for Timelab. It doesn't play well with Wine in Linux, but
that's not terribly surprising, all things considered.
Really? How so?
Wine tends to be a bit more pedantic than windows itself, but
most things, that don't go out of their way to break Wine, seem
to work... even a few viruses.
When I tried timelab, on Wine, it seemed to work ok for me.
Lady Heather also works nicely on Wine.
-Chuck Harris
Hi
One of the (many) nice thing(s) about TimeLab is that you can run multiple plots on the same data. Often
when you see wiggles in one plot, looking at another plot can help you figure out what is going on.
The “best guess” when you see wiggles on the left side, is a spur or beat note.
Bob
On May 13, 2015, at 11:11 PM, Bob Stewart bob@evoria.net wrote:
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
http://evoria.net/AE6RV/5370A/Test1.png
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
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.
Hi Bob,
The beat note idea occurred to me, as well. I see the "disturbance in the force" when using the internal clock (10811) of the 5370A, as well as when using the 10811 from the 5335A as the reference. But, if this were the case, I would think it would be a well known phenomenon. Assuming you have a 5370A/B, do you see it on your end, or do you just always use a Cs as your reference clock? I see a lot of discussion about comparing two devices, but never any discussion about the reference clock for the equipment that performs the measurement.
I suppose I could warm up my spare 34310-T and see what happens when using that as the reference.
Bob
From: Bob Camp <kb8tq@n1k.org>
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Thursday, May 14, 2015 6:06 AM
Subject: Re: [time-nuts] New 5370A
Hi
One of the (many) nice thing(s) about TimeLab is that you can run multiple plots on the same data. Often
when you see wiggles in one plot, looking at another plot can help you figure out what is going on.
The “best guess” when you see wiggles on the left side, is a spur or beat note.
Bob
On May 13, 2015, at 11:11 PM, Bob Stewart bob@evoria.net wrote:
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
http://evoria.net/AE6RV/5370A/Test1.png
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
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.
Hi Chuck,
I've renamed this thread, and expect it to die quickly. I have a bunch of things running on my Linux system, so I see a lot of interactions that probably don't happen to others. One thing is that Timelab doesn't like to come up properly under Wine. For some reason, it mostly comes up with the chart/plot covering the whole frame, rather than having the legends at the bottom and at the right.
The biggest impact to me is that it doesn't play well with MPLABX. But, I've seen other problems related to MPLABX, so suspect that the problem lies there.
Finally, I've noticed some interaction between Timelab and VirtualBox. I was running a Windows XP client, and the client got destroyed. That's never happened before. OTOH, I was trying to use a clone 82357B from the boxed client, so that may have somehow bled into Timelab. I dunno. Of course, it could have just been a random failure that appears to correlate with Timelab.
Note that I'm not shaking a finger at John. Wine is essentially the wild west at times, and this is one of those times.
Bob
From: Chuck Harris <cfharris@erols.com>
To: Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Thursday, May 14, 2015 12:21 AM
Subject: Re: [time-nuts] New 5370A
Bob Stewart wrote:
... And kudos to John for Timelab. It doesn't play well with Wine in Linux, but
that's not terribly surprising, all things considered.
Really? How so?
Wine tends to be a bit more pedantic than windows itself, but
most things, that don't go out of their way to break Wine, seem
to work... even a few viruses.
When I tried timelab, on Wine, it seemed to work ok for me.
Lady Heather also works nicely on Wine.
-Chuck Harris
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 Magnus,
Would the 5335A have the same sort of clock noise that I'm seeing on the 5370? I see it when testing on both the 5335A and the 5370A. I'm setting up a new test as follows: Internal 5370A clock, PPS from GPSDO to arm input, PRS45A to start input, and a 6 ft piece of RG-58 from start to stop input. Do you have the same sort of trace on your end when using the internal clock of whatever your test unit is?
Bob
From: Magnus Danielson <magnus@rubidium.dyndns.org>
To: time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Wednesday, May 13, 2015 11:52 PM
Subject: Re: [time-nuts] New 5370A
Hi Bob,
On 05/14/2015 05:11 AM, Bob Stewart wrote:
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
The wiggle indicates that you have some form of disturbance on top of
your signal. It could be hum, it could be the 5 MHz PWM noise that the
5370A/B clock buffer board produces. Anyway, that wiggle is typical of a
noise-source and ADEV isn't always the best tool to find out what it is.
A FFT would potentially be better for analyzing that issue.
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
Yes, it's a good tool and it's sad that it doesn't play well with Wine
in Linux. I think John pointed out that Wine did not treat a
line-drawing call correctly. Anyway, regardless of who is at fault, it's
sad they don't play well together.
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.
Hi Bob,
No, the 5335A does not have anything near the same mechanism as the 5370
in this regard.
My lab is not in order for doing such tests quickly, but let me see what
I can do about it.
In general, I rarely see that wiggle on the basic setup.
Cheers,
Magnus
On 05/14/2015 06:39 PM, Bob Stewart wrote:
Hi Magnus,
Would the 5335A have the same sort of clock noise that I'm seeing on the 5370? I see it when testing on both the 5335A and the 5370A. I'm setting up a new test as follows: Internal 5370A clock, PPS from GPSDO to arm input, PRS45A to start input, and a 6 ft piece of RG-58 from start to stop input. Do you have the same sort of trace on your end when using the internal clock of whatever your test unit is?
Bob
From: Magnus Danielson <magnus@rubidium.dyndns.org>
To: time-nuts@febo.com
Cc: magnus@rubidium.se
Sent: Wednesday, May 13, 2015 11:52 PM
Subject: Re: [time-nuts] New 5370A
Hi Bob,
On 05/14/2015 05:11 AM, Bob Stewart wrote:
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
The wiggle indicates that you have some form of disturbance on top of
your signal. It could be hum, it could be the 5 MHz PWM noise that the
5370A/B clock buffer board produces. Anyway, that wiggle is typical of a
noise-source and ADEV isn't always the best tool to find out what it is.
A FFT would potentially be better for analyzing that issue.
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
Yes, it's a good tool and it's sad that it doesn't play well with Wine
in Linux. I think John pointed out that Wine did not treat a
line-drawing call correctly. Anyway, regardless of who is at fault, it's
sad they don't play well together.
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.
Hi
I do not see it on my 5370’s but that is hardly a conclusive answer to the question. There are a number of adjustments involved in aligning the beast. There is also the variable of how far off frequency the internal standard is. I do not normally use the internal reference on any of my counters for anything other than a cleanup oscillator. I seem to have … umm …. errrr …. several … GPSDO’s and Rb’s to pick between. There is no real reason to run on the internal standard when “something better” is available.
Bob
On May 14, 2015, at 12:11 PM, Bob Stewart bob@evoria.net wrote:
Hi Bob,
The beat note idea occurred to me, as well. I see the "disturbance in the force" when using the internal clock (10811) of the 5370A, as well as when using the 10811 from the 5335A as the reference. But, if this were the case, I would think it would be a well known phenomenon. Assuming you have a 5370A/B, do you see it on your end, or do you just always use a Cs as your reference clock? I see a lot of discussion about comparing two devices, but never any discussion about the reference clock for the equipment that performs the measurement.
I suppose I could warm up my spare 34310-T and see what happens when using that as the reference.
Bob
From: Bob Camp <kb8tq@n1k.org>
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Thursday, May 14, 2015 6:06 AM
Subject: Re: [time-nuts] New 5370A
Hi
One of the (many) nice thing(s) about TimeLab is that you can run multiple plots on the same data. Often
when you see wiggles in one plot, looking at another plot can help you figure out what is going on.
The “best guess” when you see wiggles on the left side, is a spur or beat note.
Bob
On May 13, 2015, at 11:11 PM, Bob Stewart bob@evoria.net wrote:
I got a new 5370A, so of course I've been running a bunch of tests on it. In the image linked below, the start channel is my PRS-45A, the stop channel is my GPSDOe ("e" is for engine) and the ARM channel is fed by the PPS from my SSR-6Tru. As you can see in the notes, the test parameters are all the same, except for the clock source for the 5370A. As you might guess, my question is about the blue line, which is where the 5370A uses its internal 10811 as its clock reference. Is it normal for tests run like this to have the left side of the ADEV be such a "wiggle" on the internal reference? I had noticed the same thing with the 5335A I've been using up till now and had just assumed there was a problem with its 10811. The OCXO in my GSPDO is a surplus Trimble 34310-T. It's been running for several weeks, but is still in retrace.
http://evoria.net/AE6RV/5370A/Test1.png
And kudos to John for Timelab. It doesn't play well with Wine in Linux, but that's not terribly surprising, all things considered.
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 Bob,
I just got back from my son's commencement ceremony, and finally
have a little time to maybe give a few pointers.
First, safety prevents one from trying to run too many different
applications within a given wine sandbox (aka bottle). It is
best to give each application its own sandbox, and let it run on
its own... the same is true sometimes with windows... the cause
of most problems in windows is odd combinations of DLL's provided
by odd combinations of applications.
Second, if you see a completely messed up screen, it is very likely
due to using a non windows compatible font in wine. It is best
to use the real msoft true type fonts... for whatever reason (most
likely poor documentation), the clone true type fonts aren't very
good clones.
Third, running multiple virtual windows devices on the same linux
machine is simply asking for trouble.
You wouldn't consider running multiple copies of XP under XP, why
would it seem ok to run multiple XP virtualizations under linux?
Each of the virtualizations uses hooks into kernel routines, and
it is likely that they get in each other's way.
I have found timelab to work ok under wine, in its own sandbox,
with no other virtualizations running.
For what it is worth, IMHO, the best version of wine is the one
sold under the name CrossOver, by CodeWeavers. They are using a
plain version of wine, with some packaging to make the different
applications work better. They also have a package of scripts to
configure wine for a great number of different packages. The
scripts pull in any needed windows DLL's, and applications.
It's cheap to buy a years support, and they give all of their
enhancements back to the wine project.
-Chuck Harris
Bob Stewart wrote:
Hi Chuck, I've renamed this thread, and expect it to die quickly. I have a bunch
of things running on my Linux system, so I see a lot of interactions that probably
don't happen to others. One thing is that Timelab doesn't like to come up
properly under Wine. For some reason, it mostly comes up with the chart/plot
covering the whole frame, rather than having the legends at the bottom and at the
right. The biggest impact to me is that it doesn't play well with MPLABX. But,
I've seen other problems related to MPLABX, so suspect that the problem lies
there. Finally, I've noticed some interaction between Timelab and VirtualBox. I
was running a Windows XP client, and the client got destroyed. That's never
happened before. OTOH, I was trying to use a clone 82357B from the boxed client,
so that may have somehow bled into Timelab. I dunno. Of course, it could have
just been a random failure that appears to correlate with Timelab.
Note that I'm not shaking a finger at John. Wine is essentially the wild west at
times, and this is one of those times.
Bob