[time-nuts] Measuring Phase difference between different GPSDO

Bob Camp lists at rtty.us
Fri May 3 21:34:40 EDT 2013


Hi

OCXO's don't really like major temperature changes. They are much happier in steady state environments. Transient behavior may or may not correlate with steady state behavior, so trying to correct it by fiddling the oven - not a good idea. The one exception to that would be if the oven has burned out. In that case the EFC voltage should be pegged. 

-------------------

Better way to start:

Fire up the software of your choice on the PC. Let it talk to the Z3805's make sure that it's doing what it should (locking up to GPS etc). I'm guessing that's already taken care of. 

Let it run for a while and watch the EFC voltage. It'll start out moving fairly quickly. After a few days / weeks it should settle out. It should be near center scale after settling. 

While the OCXO is settling out, you can watch the PPS output and the 10 MHz. The 10 MHz should be pretty close after the first day or two. In this case pretty close is < 0.00001 Hz on the 10 MHz. Pretty close on the pps should be < 20 ns. If they are not, the disciplining part of the unit is broke.

Depending on what you do / don't find there are lots of different directions to head. 

Bob

On May 3, 2013, at 5:04 PM, "Mark C. Stephens" <marks at non-stop.com.au> wrote:

> Actually Bob, What put me onto this was one of the Z3805A, the PU decreases dramatically when a cooling fan is pointed at it.
> 
> The Second one, is the opposite, positioned near the hot air exhaust of a server gets the best PU.
> 
> I have 3 other Z3805A which I have yet to play with.
> 
> Am I onto something or clutching at straws?
> 
> 
> mark
> 
> -----Original Message-----
> From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] On Behalf Of Bob Camp
> Sent: Saturday, 4 May 2013 3:01 AM
> To: Discussion of precise time and frequency measurement
> Subject: Re: [time-nuts] Measuring Phase difference between different GPSDO
> 
> Hi
> 
> Sounds like you are up and running. TimeLab can be set for any length run you wish. Three to five minutes is a common one. I have a 30 day run going at the moment. Even while the run is going, you can switch between displays to see what's happening. 
> 
> Unless you have some very unusual OCXO's in your 3805's fiddling with the oven temperature isn't going to help things much. If they are SC based, there may not be a turn over temperature. Often the set point is derived either from crystal data or from wide range temp run performance. 
> 
> Like any piece of surplus, you can indeed have something broken somewhere. In that case, repair is a good idea. Unless you blew the thermistor in the OCXO, it's an unlikely repair though.
> 
> Bob
> 
> 
> On May 3, 2013, at 12:48 PM, "Mark C. Stephens" <marks at non-stop.com.au> wrote:
> 
>> 
>> 
>> Am on the way to doing my first 'phase difference' with timelab - 
>> takes an hour :)
>> 
>> This is good stuff, everything went so smoothly, from installing the OS to using Timelab.
>> 
>> Now I have to figure out how to measure and set the xtal turning temperature for this bunch of Z3805A's I have here.
>> 
>> I have a Z3816A that is super stable. SYST:STAT? usually reports 0.0us 
>> for PU, sometimes it will go out as far as 0.2us, but very rarely :p
>> 
>> So I am planning on using that guy as my reference to tune the Z3805A's.
>> 
>> How does that sound?
>> 
>> 
>> Many thanks,
>> mark
>> 
>> 
>> -----Original Message-----
>> From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] 
>> On Behalf Of Bob Camp
>> Sent: Saturday, 4 May 2013 1:09 AM
>> To: 'Discussion of precise time and frequency measurement'
>> Subject: Re: [time-nuts] Measuring Phase difference between different 
>> GPSDO
>> 
>> Hi
>> 
>> Grab an old garage sale category PC, blow the dust out of it and fire up the GPIB. Once it's all working next steps:
>> 
>> Check out TimeLab and EZGpib. Both are pretty good at talking to "stuff".
>> TimeLab does support the 5370, not sure about the other programs.
>> 
>> If you have two GPSDO's with pps outputs, feed one into each channel on the 5370. Put it in start / stop mode and you should see the difference between the two. Since they can be either ahead or behind each other things can get a little messy. 
>> 
>> Next, go to the software that runs your GPSDO and look for the "cable delay"
>> setting. Adjust it on one GPSDO so it's always early (or late) by a couple hundred nanoseconds. Now fiddle the cables to the counter so it reads a couple hundred nanoseconds and not almost a full second :).
>> 
>> At this point you are reading frequency to a very high resolution. The GPSDO's likely are only good to a couple of nanoseconds at 1 second. The counter should be good to ~0.02 ns.
>> 
>> Run like this for a few minutes and see what you get.
>> 
>> -----------
>> 
>> Now that you have conquered that approach it's time to tear it all apart and try something new.
>> 
>> Feed the 1 pps into one input of the counter (just like before). Hit the other input with 10 MHz. You now will get a reading in the 0 to 100 ns range. Each time a cycle slips past the 1 pps the reading will skip from 99.xxx to 0.xxx ns. That's called a phase wrap around. Click the appropriate boxes on which ever software you are running and it should take care of the wrap.
>> 
>> --------
>> 
>> Either way the software should show you frequency, and frequency stability. 
>> 
>> Bob
>> 
>> 
>> -----Original Message-----
>> From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] 
>> On Behalf Of Mark C. Stephens
>> Sent: Friday, May 03, 2013 10:07 AM
>> To: Discussion of precise time and frequency measurement
>> Subject: Re: [time-nuts] Measuring Phase diference between different 
>> GPSDO
>> 
>> Hey Bob, I was thinking the exact same thought, and I found a NI PCI-GPIB card in a PC with a dead HDD.
>> These appear to be 5V PCI cards so I can't stick it in a nice rack mount server with raid as they all seem to be 3.3v PCI.
>> So I put a new HDD in the PC and I have stuck WinXP on it.
>> Now I am awaiting further orders!  ;)
>> 
>> mark
>> 
>> -----Original Message-----
>> From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] 
>> On Behalf Of Bob Camp
>> Sent: Friday, 3 May 2013 9:21 PM
>> To: Discussion of precise time and frequency measurement
>> Subject: Re: [time-nuts] Measuring Phase diference between different 
>> GPSDO
>> 
>> Hi
>> 
>> If you have a pile of GPIB cables, you may also have a GPIB card on a PC.
>> You will have a lot more luck finding free software to read and log the data from the 5370 than from the 3575. As mentioned before, the 5370 will be much more accurate / higher resolution in this application. 
>> 
>> Bob
>> 
>> On May 3, 2013, at 12:28 AM, Mark C. Stephens <marks at non-stop.com.au> wrote:
>> 
>>> According to http://www.realhamradio.com/z3801a-turning-point.htm 
>>> they use
>> a HP 3575A phase meter to perform the measurement.
>>> Or perhaps I have misinterpreted the whole thing?
>>> 
>>> 
>>> I do have a temperature controlled workshop that is always 24 degrees 
>>> so
>> hopefully thermal drift won't be too much of an issue.
>>> 
>>> 
>>> mark
>>> 
>>> -----Original Message-----
>>> From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com]
>>> On Behalf Of lists at lazygranch.com
>>> Sent: Friday, 3 May 2013 12:12 PM
>>> To: Discussion of precise time and frequency measurement
>>> Subject: Re: [time-nuts] Measuring Phase diference between different 
>>> GPSDO
>>> 
>>> Meditating on this a bit, I assume in a strict sense, you can only
>> consider GPSDOs phase locked if they are disciplined from the same GPS.
>>> 
>>> Or is this being pedantic?
>>> 
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at 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 at 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 at 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 at 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 at 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 at 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 at 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 at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.



More information about the time-nuts mailing list