[time-nuts] Question for any HP105 owners

[Tom Van Baak]

Mark,

One more comment on my 1e-12 example. The measurements
that you obtain fix the limits of how bad either oscillator can be;
but it doesn't prove how good either one is.

In one extreme the oscillators could have near equal stability,
say 7e-13 each. As mentioned earlier, that would give you a
measured value of 1e-12 (one for REF, the other for DUT).

But other combinations can also yield a rms measurement
result of 1e-12. For example, 6e-13 and 8e-13, or 2e-13 and
9.8e-13, or 1e-13 and 9.95e-13. So from this you can't really
tell how good one of the oscillators is; all you can be sure of
is that neither one is worse than 1e-12.

It's tempting to think of both oscillators being equal as some
sort of "best case", that is, when you compare them you get
1e-12 and that means each is actually 7e-13. Sounds good.

But I'm sure you agree it would be even better if you later
discovered that the reason your measurement result is 1e-12
was that one oscillator was 1e-12 and the other was 1e-13!

So use the sqrt(2) trick carefully.

/tvb

----- Original Message ----- 
From: "Mark Spencer" <mspencer12345 at yahoo.ca>
To: "Tom Van Baak" <tvb at leapsecond.com>; "Discussion of precise time and frequency measurement" <time-nuts at febo.com>
Sent: Tuesday, August 30, 2011 12:29 PM
Subject: Re: [time-nuts] Question for any HP105 owners


Thanks Tom that makes sense.    (In this case I'm fairly sure that the FTS 1050 is at least meeting it's 100 sec ADEV 
spec of 1E-12 I have had results in the 13's on occasion when looking at other oscilators, and I leave it running in a 
fairly stable enviornment.)

I appriceate the insight.

Regards
Mark Spencer