[time-nuts] FE-5680A arrived
mike cook
michael.cook at sfr.fr
Wed Jan 11 21:47:10 UTC 2012
Le 11/01/2012 19:36, mike cook a écrit :
> Le 11/01/2012 19:24, Chuck Forsberg WA7KGX N2469R a écrit :
>> The FE-5680A 1pps is not sync'd to UTC.
>> According to my calculations, it would take several hours
>> to walk the 1pps to UTC given the 3.8e-5 adjustment
>> range via the RS232 input. A thunderbolt, an arduino,
>> and patience.
> You don't need to take the walk. Just fudge the ATOM driver time1
> parameter after scoping the offset between GPS PPS and 5680A PPS
> signals. You would need to reset it on a Rb power cycle though.
>
just thought of something else which worked for me and may be of use.
If you don't have a scope or TIC to get the Rbs PPS offset from GPS,
you can let NTP do it for you. You need a stable reference such as GPS
to be set up as a prefered server, then set up the atom driver for the
5680A with no offset and prevent selection with "noselect" parameter and
or push the stratum from 0 to 8. Once NTP has been running a while, the
offset of the DUTs PPS will be seen in the ntpq -pn output. You can plug
that in as the DUTs fudge time1 parameter and restart ntpd. You may need
to iterate.
ex.
server 127.127.22.1 minpoll 4 maxpoll 4 noselect
fudge 127.127.22.1 stratum 8 refid PPS1 time1 0.000000 flag3 0 flag2
1 # enable PPS module
remote refid st t when poll reach delay offset
jitter
==============================================================================
127.127.22.1 .PPS1. 8 l 9 16 377 0.000
-1.268 0.093
*145.238.203.14 .TS-3. 1 u 67 64 377 34.473 1.572
0.399
remote refid st t when poll reach delay offset
jitter
==============================================================================
127.127.22.1 .PPS1. 8 l 11 16 377 0.000
-1.528 0.038
*145.238.203.14 .TS-3. 1 u 19 64 377 33.300 0.217
0.750
ntpq -pn only gives usec resolution displays but ntpd calculates in
nanosecs. You have to have logging enabled and look at the peerstat data
if you want that.
More information about the Time-nuts_lists.febo.com
mailing list