[time-nuts] HP5065a cfield issue

[Poul-Henning Kamp]

--------
paul swed writes:

>Bob mentioned flipping the synthesizer switch. Heck they have been sitting
>at 8714 (Also on the tube label) for 30 years. I did  and it did not fix
>the problem. But I went through the switch seeing how each section behaved.
>By dropping to 8614 [...]

Please notice that the settings are not sequential, you should
use the settings in the sequence they are listed in the manual.

> Will use a 5371b counter later tonight and get serious.

Hook it up to the synth output, and use histogram mode:

One failure mode of the synth is "hunting" between two frequencies
because of meta-stability/failure of the unobtanium delay chip in
the divider chain.

One of my HP5065s did that:

	http://phk.freebsd.dk/hacks/HP5065A/20160408_synth/

I replaced the entire divider chain with a LPC810 ARM chip where
I programmed the PWM output to output the correct frequency and
duty-cycle.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.