[time-nuts] another Ebay mixup, 5370

Magnus Danielson magnus at rubidium.dyndns.org
Sun Jun 10 09:48:11 EDT 2007

From: Dr Bruce Griffiths <bruce.griffiths at xtra.co.nz>
Subject: Re: [time-nuts] another Ebay mixup, 5370
Date: Mon, 11 Jun 2007 01:32:51 +1200
Message-ID: <466BFD83.9020109 at xtra.co.nz>

Tom, Bruce,

> >> The last time I studied the manual, I remember being convinced that the 5370
> >> would be relatively immune to clock-correlation effects if the interpolators
> >> were set up properly.
> >>     
> >
> >   
> Wishful thinking, the designers don't believe this so why should we?
> I have a copy of a letter from David Chu detailing the then known 
> sources of differential nonlinearity in the 5370A.
> These causes are principally crosstalk between the START and STOP 
> channels, and interaction between the mixers and the phase locked 
> oscillators.

The crosstalk is there, and will affect your data. The benefit of using the
same clock as the internal measurements is made out of is that cross-
correlation with the internal reference remain the same. This can also be
viewed as a problem depending on what you do, so then a different clock is
needed such that such effects can be averaged out. Depends on your needs.
Regardless of which, you need to understand how different crosstalk effects
intermingles with your measurement so that you can either avoid it or remove
the major effect from it. This is the key to use an instrument beyond its
specifications. Learn the quirks and learn to handle them for your particular
measurement. You milage *WILL* vary.

We have been over this a number of times.

An easy exercise is to hook either START or STOP to the internal reference and
then have another 10 MHz source with a slight offset running on the other.
Record the data until it clearly loops (it is apparent). Then use a T and equal
length cables to measure the zero time and you can relate data to TIE for a
certain TI value. If you have the calibrator box you can do it even better.
If you swap the cables you can cancel out their length difference.


More information about the time-nuts mailing list