[ntp:questions] fudge time1 for gps-18x-LVC?

Dave Hart davehart at gmail.com
Sun Feb 7 16:12:31 UTC 2010


On Feb 7, 9:35 UTC, "David J Taylor" wrote:
> My (rather limited) understanding was that the
> atom driver was simply a filter on an existing PPS signal, and therefore
> you would need a driver which knew about the LPT port and could detect and
> use its PPS signal.  Which driver does that I don't know, as I've never
> used one.

There used to be a grain of truth in what you say, but with the
current code, the grain is gone.

It used to be the case that drivers that implemented their own PPSAPI
handling, including the NMEA driver (20), relied on code in the ATOM
driver (22) to filter the PPS samples over a polling interval into a
single offset.  Last April, Dr. Mills relocated that code to
ntp_refclock.c in new common PPSAPI handling code used by several
drivers, including the Spectracom/WWVB driver, NMEA driver, ATOM
driver, and to some extent, the parse drivers.

The driver which supports PPS independent of any other timecode/
refclock is the ATOM driver, number 22.  The PPSAPI code in 4.2.6's
ATOM and NMEA drivers is essentially the same, the difference is
whether it's used independently or associated with a NMEA source.
Used independently, there must be at least one other source marked
prefer.

Cheers,
Dave Hart




More information about the questions mailing list