[time-nuts] 5061 Cs frequency shift with long time constant?

Bill Hawkins bill at iaxs.net
Tue Apr 5 14:51:51 EDT 2005

If a capacitor is used for integration then there should only be
a phase offset unless it is really leaky. The integrator is inside
of a feedback loop, so that internal effects should be able to be
removed - except that capacitor leakage causes an offset that
must be compensated by more phase error.

Please check the schematic. I thought long-term integration could
only be done with a digital integrator.

Bill Hawkins

-----Original Message-----
From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com]On
Behalf Of Hubert v. Bonhorst
Sent: Tuesday, April 05, 2005 1:32 PM
To: 'Discussion of precise time and frequency measurement'
Subject: AW: [time-nuts] 5061 Cs frequency shift with long time

there is some leak-current or offset in the integrating amplifier. Suspect
the amplifier itself as well. Try to do the integrating amplifier adjustment
procedure. Eventually modulator phase needs a little bit adjustment. Be
aware that a instrument connected to control output could cause a similar
effect. If you consider the specification. The 5061A is within limits.
Have fun

-----Ursprüngliche Nachricht-----
Von: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] Im
Auftrag von Poul-Henning Kamp
Gesendet: Dienstag, 5. April 2005 18:50
An: Discussion of precise time and frequency measurement
Betreff: Re: [time-nuts] 5061 Cs frequency shift with long time constant?

In message <4252B46E.3010509 at pacific.net>, Brooke Clarke writes:

>> I got my 5061A trimmed to the point where a 5 day run against GPS
>> showed an offset of less than 1x10e-13.  I then switched the unit into
>> the long time constant mode and now see that the offset (over 36
>> hours) is -1.7x10e-12.  Nothing else in the setup has changed.

Yes, very likely.  The leak-current in the integrating capacitor cannot
be fully ignored at this level.

Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.

More information about the time-nuts mailing list