[time-nuts] Trimble Resolution T - PPS offset and stability

Bob Camp lists at rtty.us
Tue Sep 3 00:10:12 UTC 2013


Hi

Ok, well if you are doing common view between Time Nuts, then it's highly likely that an accurate / stable 10 MHz is available. Shove this into the clock port on the STM32F4 and have it time stamp accurately. You could also do an analog TDC for a couple bucks if you wanted to do things a bit differently. More or less:

CPLD for $3 or so for the counters / timers / logic.
Cheap MCU from whom ever for < $5
Analog TDC outside the micro $3, inside the micro - $0.25 (once you have the CPLD).

With a 100 ns timebase in the CPLD (10 MHz you already have) the TDC would run from 100 to 200 ns. If that maps to 10 to 12 bits, you get maybe 1,000 useful buckets. That will get you down to the vicinity of 200 ps. 

Bob

On Sep 2, 2013, at 7:12 PM, Lachlan Gunn <lachlan at twopif.net> wrote:

> It sounds like an experiment is in order, then.
> 
> To give some background, around six months ago I posted on here about the
> idea of running an amateur common-view GPS network over the internet.  This
> is all well and  good, but if I'm to have anyone else to compare with then I
> will need to bring the hardware down to $20 from Farnell/Digikey/etc. and
> five minutes of work.
> 
> At least initially, this will mean a microcontroller rather than an FPGA to
> make the comparison.  I've been using a $15 STM32F4 devboard as a DSP
> platform, but it has timers that can run at 168MHz, and can use them for
> event timestamping.  Not as fast as I've done with FPGAs, but it could
> certainly be worse.  Using something like this means that the timebase
> stability will be awful, and so I'd rather use the PPS offset to avoid
> degradation as a result if possible.
> 
> Thanks,
> Lachlan
> 
> -----Original Message-----
> From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] On
> Behalf Of Bob Camp
> Sent: Tuesday, 3 September 2013 7:54 AM
> To: Discussion of precise time and frequency measurement
> Subject: Re: [time-nuts] Trimble Resolution T - PPS offset and stability
> 
> Hi
> 
> On all the GPS's I have tried it on, shifting the PPS has no real impact on
> stability. A few things to consider:
> 
> Normally the shift is a few hundred ns either way The shift process is
> always in steps of the main clock (100 ns for 10 MHz) GPS by it's self
> bounces around a bit.
> 
> If you are talking about a shift of a big fraction of a second (and it
> sounds like you are) then the stability of the GPS's local clock could come
> into play. On something like a TBolt that's not going to matter. On a TCXO
> based gizmo that is only corrected to 1.0x10^-7 you could get an extra 50 ns
> of error at a half second offset. Weather you see that on this or that GPS
> depends a lot on who wrote the firmware and what they worried about when
> they did. 
> 
> The better alternative is to use a counter with a reasonable time base to
> look at the difference between pps signals. If the counter has an OCXO time
> base and it's properly calibrated you are about 10 to 100X better off than
> the 50 ns in the example above. 
> 
> Bob
> 
> On Sep 2, 2013, at 6:05 PM, Lachlan Gunn <lachlan at twopif.net> wrote:
> 
>> Hello.
>> 
>> 
>> 
>> Has anyone here tried varying the PPS offset on a ResT (or UT+ or any 
>> other GPS receiver for that matter) and measuring the resulting stability?
>> 
>> 
>> 
>> I ask because my Rb has only a 1PPS output, and while I have been able 
>> to get at one of its internal HF signals, would like to see what I can 
>> do with just 1PPS.  The obvious problem with doing this is that I will 
>> need to align the PPS outputs together to get reasonable accuracy, but 
>> I worry that a large offset in the GPS receiver will degrade stability 
>> as the pulse moves away from the relevant packet in the GPS signal.
>> 
>> 
>> 
>> Am I being over-cautious?
>> 
>> 
>> 
>> Thanks,
>> 
>> Lachlan
>> 
>> _______________________________________________
>> 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_lists.febo.com mailing list