[time-nuts] Differences / Jitter Between Frequency Standard Sources
Tom Van Baak
tvb at LeapSecond.com
Thu Nov 24 17:51:46 EST 2011
> I have a HP 5385A frequency counter, which I am using with an external
> standard. If I try to compare one 10 MHz standard against another,
> using one as the external standard and the other as the device being
> measured, the readout fluctuates between 10.0000000 and 9.9999999. This
> happens when I am comparing a Z3801A and the Trimble look-a-like using
> the same GPS antenna and identical cables (including length) between the
> GPS antenna splitter and the two units. I have also used identical
> cables between the 10 MHz outputs and the counters. Hence, cable
> variations shouldn't be the issue.
> The same thing happens comparing the Timesource 2700 against either of
> the GPS standards.
> Is this attributable to the different standards or the frequency
> counter? I haven't yet set up a dual trace scope to compare the various
> sources directly, but that will be my next step.
This is normal (for both gated and reciprocal counters). Never trust
a frequency counter that gives you the same number every time.
In fact it is very advantages. If you log hundreds or thousands of
these frequency readings and compute the mean you will get a
more precise value for your long-term average frequency. It's
unlikely you will get a 50/50 split between the two readings; the
actual ratio between them allows you to "interpolate" somewhat
between 9 and 0 in the last digit. Try it.
More information about the time-nuts