[ntp:questions] Fwd: Re: NetBSD GPS/PPS using 4.2.6p3

unruh unruh at wormhole.physics.ubc.ca
Thu Aug 25 14:50:21 UTC 2011

On 2011-08-25, David Lord <snews at lordynet.org> wrote:
> unruh wrote:
>>> even considering the PPS and then the system clock has to
>>> be within a millisec of the PPS. Using one of your other
>> Are you sure? That seems a very verysilly requirement to put on the time
>> accuracy of the PPS. All that should be required is that the pulse be
>> within .5 sec of the "true time" in order to establish the "seconds" for
>> pps. 
>  From my reading of the docs the ntp time has to be within
> about 1 millisec of the pps before pps will sync.

Where do you read that? 

> A pps pulse within .5 sec of the "true time" doesn't make

Sorry, I meant that the time stamp on the pulse be within that (ie, the
system time be within that) I believe ntp actually requires that the
system time be withing .4 sec of true time. 

> any sense to me. Perhaps you meant the NMEA sentence used
> needed to be within .5 sec but again that doesn't make a
> lot of sense in practice as all three different gps units
> I've used each have large variations in NMEA time so
> require both fudge factor and increase in the range of
> values that will be accepted.

Nope, I meant that system time should be within .4 sec. That will not
matter if you use gpsd since that should accurately get the time of the
pps pulse from the nmea of the gps unit (assuming it is not broken like
some of the garmin 18x are).

More information about the questions mailing list