[time-nuts] Z3801 saga

[Magnus Danielson]

From: Magnus Danielson <cfmd at bredband.net>
Subject: [time-nuts] Z3801 saga
Date: Wed, 26 Jul 2006 01:32:42 +0200 (CEST)
Message-ID: <20060726.013242.-1016356108.cfmd at bredband.net>

> Fellow timenuts,
> 
> The automation fell in my lap today. The Allan deviation plot show clear
> evidence of CW modulation (traditional ripple). Turing over to frequency view
> there it is. Actually, it has both varying amplitude and frequency. As
> frequency lowers, so does the amplitude in frequency deviation. When it hits
> around 450 mHz on the 10 MHz, the Z3801A goes into holdover.
> 
> Now, it seems that there is some source of pumping up and down. I am suspecting
> something like the oven control pumping. The longer time it spends the longer
> the temperature goes up and down and thus modulate the frequency more and more.
> 
> Somebody sitting there with the outer and inner oven control schematics. I've
> only seen parts of the inner one.
> 
> Does this ring a bell somewhere?

Eh... forget THAT one. THAT was a result of a measuring mistake. Somewhere down
the line I got into running the internal timebase which isn't fantastic as
default, but I was beleiving I was running fancier... (Cs-beam vs.
uncompensated crystal)

Now it looks cooler. Much cooler. It still goes out of lock and it goes way out
of range. Last time I looked it was dangerously close to the 1 us TI error
limit at which time it goes into holdover and reallign itself. I'm a bit
unhappy about that.

Anyway, I keep it heated while I'm gone to see if things improve just by
staying tuned.

It's a pitty I can't receive more channels on this one, I have some 12
satelites above 10 degree threshold. Good signal on the 6 I'm tracking.

Cheers,
Magnus