[time-nuts] Z3801 lockin behaviour

[Magnus Danielson]

From: John Ackermann N8UR <jra at febo.com>
Subject: Re: [time-nuts] Z3801 lockin behaviour
Date: Mon, 24 Jul 2006 13:18:06 -0400
Message-ID: <44C500CE.4070106 at febo.com>

> Magnus, I like Tom's idea of plotting the 10MHz with no GPS signal, to
> see if it's the oscillator or not, but I would go one step further and
> disconnect the EFC input from anything to rule out any noise in the
> control circuitry.

Actually, that is not the reason for disconnecting the EFC input, the real
reason is that the EFC does not stay put just because of holdover since it has
the HP SmartClock technology. That means that it will adjust the EFC as a
result of temperature and supply rail voltages.

It is during holdover that the HP SmartClock technology is supposed to operate
actively, while it is in training during GPS tracking operation.

Never the less, it is a good idea to disconnect it.

I have now heated up my two other 10811s as well as another crystal oscillator,
and they will be heater for over an hour before I measure them. There will thus
be no less than 5 different references available. I will use two different
counters for my measurements. One is a HP 5372A (200 ps single-shot) and the
other is a Pendulum CNT-90 (which I have borrowed, claimed to be 100 ps single-
shot, but that is conservative, also, it is sligthly hacked for improved
perfomance). I don't have a HP5370 or... yes I do... a HP 5370C - which is my
scanner! Considering of getting a HP 5370 just for the sake of completeness.

Anyway, I have had two good tips comming from Fred, one relating to the oven
turnover point issue and another relating to the EFC filter circuit. Both are
highly interesting in my book. Disconnecting the EFC will help me to separate
out between which to check first!

Cheers,
Magnus