[time-nuts] Progress of my HP Z3801A -- Ooops!

Magnus Danielson cfmd at bredband.net
Wed Aug 30 17:19:29 EDT 2006


From: John Ackermann N8UR <jra at febo.com>
Subject: Re: [time-nuts] Progress of my HP Z3801A -- Ooops!
Date: Wed, 30 Aug 2006 16:30:54 -0400
Message-ID: <44F5F57E.1030200 at febo.com>

> Magnus Danielson wrote:
> 
> > Yes, there is a time-difference between the reported and the time of the
> > machine. The machine has NTP from 4 known good public Stratum 1 NTP servers.
> > the 11-13 s time-difference is "interesting".
> 
> Might your Z3801A be set to GPS time rather than UTC?  That would 
> account for (I think) 13 seconds (or is it 14 now that we've had a 
> leapsecond?).

That's it! Infact, it is apparent if one looks on the webpage:

Receiver Time: 2006-08-30T20:26:30.0 GPS
Local Time: 2006-08-30T22:26:19

Local Time is UTC + 2h, so there you have 2006-08-30T20:26:19 actually, and
infact this shows that ISO2 isn't following the ISO format.

Now, with 14 seconds of GPS/UTC offset the receiver Time should be
2006-08-T20:26:16 and the remaining difference of 3 seconds is maybe due to
the report-time???

I think it is much better to do the :SYST:TIME? to get the time, while all the
other stuff very well may come out of the :SYST:STAT? report and parsing.

Ideally it should read out the UTC-GPS time-difference and compensate with that
OR set it to UTC. Yes, I know I can add commands to the end of my conf-file.
UTC is better, since then it would handle leap-seconds +/- too early leap-
second announcements and the leap-second bug.

Cheers,
Magnus



More information about the time-nuts mailing list