[time-nuts] Re: Temperature effect on delay of FatPPS :)

[John Ackermann N8UR]

The FatPPS should be pretty stable when you trigger off the leading edge 
of the output, since the delays are in silicon.  But the trailing 
(usually falling) edge of the output is a different matter as the pulse 
width is controlled by a simple RC circuit and is likely to wander all 
over the place.

John
----

On 3/29/22 05:10, Hal Murray wrote:
> I'm putting some data collection toys back together.
> 
> I've got a TAPR TICC watching a couple of PPS signals.
> 
> The clock for the TICC comes from a HP 5334B with the good crystal option.
> It's not right-on in frequency, but there is no control voltage that might be
> wiggling around so it should be stable.
> 
> This setup is half backwards.  I'm using the PPS to calibrate the frequency of
> the TICC reference clock, then assuming the reference clock is short term
> stable and using it to examine the short term characteristics of the PPS
> signals.
> 
> The green line is the PPS signal from a Sure demo board with a SKG16B GPS chip.
>    http://www.leapsecond.com/pages/MG1613S/
> 
> The wander in the green line is a combination of warmup and temperature.  The
> 5334B was off for several months.  This graph starts on day 18.  The wiggles
> are easy to see on an expanded scale but that blows the FatPPS signal off
> scale.  The daily temperature swings are roughly the same size as the daily
> drift.  (Another graph for another message...)
> 
> The PPS signals are connected via clip leads to handy places.  One of those
> was after a FatPPS.  That signal is upside down so the TICC is triggering on
> the trailing edge.  No problem, I thought, I can just fix that up when
> graphing things.  :)
> 
> After I saw the initial results, I put a temperature probe on the FatPPS.  The
> spike at 18 hours is the direct morning sun peeking through the houses and
> trees across the street.
> 
> 
> 
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe send an email to time-nuts-leave at lists.febo.com
> To unsubscribe, go to and follow the instructions there.