[time-nuts] Datum 1000B troubleshooting

Eugene Timing timing.fanatic at gmail.com
Sun Feb 4 14:21:04 EST 2018


Hi all,

I'm still relatively new in this area but I'm trying to troubleshoot an
issue with a Datum 1000B I picked up last year.  This unit is the 10MHz
flavor with 4 outputs.  Its issues, I believe, is after 24 hours of
warm-up, the oscillator:
- ...has a flat-out dead output for one of the outputs, other 3 outputs work
- ...exhibits an unstable frequency output, from all working ports

I own only a modest set of equipment (Racal 1992 and a Thunderbolt I bought
from the group-buy years ago) for frequency testing, but my Racal is
showing me the frequency is fluctuating between just under and over 10MHz.
Someone on this list previously coined the term "poor-man's spectrum
analyzer" method with a Thunderbolt, and that's what I did - I stuck the
oscillator output to my modified external-oscillator fed Thunderbolt.  This
Thunderbolt has been used to discipline a variety of oscillators (10811A/D
and 60111s, Morion 89A) and generally shows 30-50ppt with the noisy GPS
reference.  With the 1000B it's showing about a 500-1000ppt fluctuation
with GPS.  While I know GPS is generally noisy, I'm pretty sure the 1000B
figure isn't normal.

My test setup is +24V connected to all powered inputs and the EFC input is
grounded.  The trim pot on the oscillator face does seem to be able put the
oscillator on-frequency.

I'm hoping this list can tell me what common issues this old-ish unit
generally exhibits in terms of instability before I deep dive into its
innards, but here are my thoughts:
- Large 24V supply ripples
- Bad oscillator power regulator(?)
- Failed oven heater controller to heat crystal to its turn-over point?

I don't think it's any of the output buffer circuits since this noise
appears on all working outputs...

The unit starts up with a current consumption of 410mA cold, then drops to
about 80mA @ 24V after fully warmed up.

Any suggestions/pointers from this list would be welcome!

-Eugene


More information about the time-nuts mailing list