[ntp:questions] Linux NTPd using a older Trimble Thunderbolt GPS Receiver

William Unruh unruh at invalid.ca
Tue Jan 20 18:18:32 UTC 2015


On 2015-01-20, George Ross <gdmr at inf.ed.ac.uk> wrote:
> --===============2288611982837908707==
> Content-Type: multipart/signed; boundary="==_Exmh_1421754685_7720P";
> 	 micalg=pgp-sha1; protocol="application/pgp-signature"
> Content-Transfer-Encoding: 7bit
>
> --==_Exmh_1421754685_7720P
> Content-Type: text/plain; charset=us-ascii
>
>> ... Presumably PPS
>> was ignored because the event based timing packets yield reliable
>> sub-millisecond offsets.  The driver and document should be brought into
>> the PPS era and be renamed the TSIP refclock rather than Palisades.
>>
>> Palisades/NMEA + ATOM is the way to use these receivers.
>
> From the Acutime 2000 user guide: "The time tag provides a resolution of
> 320ns ...".  Is PPS going to be sufficiently better that it would outweigh
> the additional setup complexity?

??? The question is not what the resolution of the time tag is. The
question is how accurately you can get that time into your computer.
Certainly sending over say a serial line (each bit of which takes 100
of microseconds, with fluctuations) making it impossible to get any kind
of accuracy into the computer. The PPS triggers an interrupt which can
be handled in the microsecond range of times. 



More information about the questions mailing list