[time-nuts] PPS for NTP Server - How Close Is "Good Enough"?
chris at chriscaudle.org
Wed Jun 10 13:35:28 EDT 2015
On Tue, June 9, 2015 11:30 pm, Ed Armstrong wrote:
> OK, here is the actual question. Do you think it is OK to consider a
> pulse which arise 250 ns early to be close enough?
It is only 250ns early relative to the even second output. What is the
even second output referenced to?
What most people care about is relation to established UTC, and that you
don't know for either pulse. You did not mention measuring the length of
the antenna cable and adjusting for that. That will probably be a few tens
of nanoseconds. There could be other processing delays or propagation
delays in the signal path. Without accounting for all of those, and
knowing whether the firmware in the GPSDO takes those into account, you
don't really know whether the PPS is early, or the PP2S is late relative
to UTC (and assuming you have accounted for antenna delay).
More information about the time-nuts