[time-nuts] PIC Divider Performance (was Re: time-nuts Digest, Vol 56, Issue 71)
bruce.griffiths at xtra.co.nz
Tue Mar 31 03:03:25 UTC 2009
Tom Van Baak wrote:
>> Over 1000 measurements at 1 PPS, I got a standard deviation of 46.9 ps.
>> The delta between minimum and maximum readings was 330 ps. Since we
>> were testing two independent dividers, I suppose you could divide the
>> standard deviation by the square root of 2, which gives about 33 ps.
>> I then did a test with a common PPS signal driving both inputs to the
>> 5370B, with the stop signal run through a 4 meter cable for delay. This
>> should show the noise floor of the TIC. The results for 1000 samples
>> there were 23.0 ps standard deviation, and delta between min and max
>> readings was 140 ps.
>> Therefore, the PIC divider is a bit above the 5370B noise floor, but not
> Thanks for doing this. That technique, and your results,
> sound similar to what I did when I tested the divider. Cool.
> I suspect the 5370 standard deviation will even decrease
> further when the two dividers are sync'ed closer. Part of
> what you're seeing is jitter in the 5370 timebase, I think.
> One thing I'd like to point out -- if one is using a divider in
> order to produce a 1PPS in order to compare against a
> reference 1PPS using a ns or sub-ns TIC then the divider
> should be better than the TIC or at least on par with it, but
> it doesn't have to be perfect (if such a thing exists in the
> world of timing).
> Since I do most of my 1PPS timing with TICs like a sr620
> (25 ps) or hp 53131 (750 ps) or hp 53132 (150 ps), the
> divider noise is minimal to nothing compared to the jitter or
> resolution of the counter itself. This is even more so when
> the reference 1PPS comes from a GPS engine (as is the
> case when a PIC divider is used in a GPSDO).
> 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.
If the divider output spectrum were examined then the 100KHz output will
cause spurs about -120dBc at 100KHz from the carrier (assuming ~10ps of
phase modulation due to ground bounce etc).
Detecting such spurs may be a little difficult particularly if such
phase modulation is only a picosec or 2.
The amount of such phase modulation depends on the load of the output
causing such modulation so heavily loading such outputs would improve
the likelihood of detection.
Other divider outputs other than 100KHz will also contribute phase
modulation spurs of a similar magnitude even closer to the carrier.
Since the divider outputs approximate a square wave the spectrum will be
even more complex than stated above.
More information about the time-nuts