[time-nuts] Why is holdover LED on HP 58503A not lit

Charles Steinmetz csteinmetz at yandex.com
Mon Dec 19 21:30:17 EST 2016

David wrote:

> This does appear now to be working - see below. I'm not exactly sure what
> was the secret to getting this to work, but the above procedure seemed ok.

Time.  It is not at all unusual for a 58503A that has no almanac (or an 
out-of-date almanac) to take several days to achieve disciplined 
operation (and much longer for the OCXO to fully stabilize).  You may 
get lucky and find that it happens faster, but don't count on it.  FWIW, 
HP GPSDOs often take longer than others to initialize.

Everyone, please remember -- GPSDOs can take significant time to 
re-initialize.  No need to post about what may be wrong until at least 3 
days of continuous running after powering on.  Chill.

Same with respect to crystal oscillators settling to low drift rates. 
No need to worry about drift (retrace) for at least a month (sometimes 6 
or even 12 or 24 months) of continuous running after powering on.

Honestly, the best thing to do is turn it on and forget about it for 3 
days (GPSDO) or 1 month (crystal oscillator).  DO NOT observe it (log 
it, if you feel the need, and don't look at the log until after the 3/30 

Precise time and frequency require patience!

Best regards,


More information about the time-nuts mailing list