[time-nuts] Question on crystal jumps

SAIDJACK at aol.com SAIDJACK at aol.com
Thu Oct 23 16:25:44 UTC 2008


Hi Bob,
 
they should and do, and if it jumps the Rb will pull-in the OCXO  frequency 
correcting for the error. This pull in time is determined by the loop  time 
(cut-off frequency) of the Rb PLL loop. I expect this loop time to be much  
shorter than a GPSDO's loop time, so the effect is probably much less visible.  One 
should still be able to see it as a distinct offset when plotting the  EFC 
control voltage of the OCXO.
 
Considering that jumps often hit 1ppb offset, I am surprised how many folks  
on this forum have discussed that their oscillators jump, and how little  
research actually has gone into figuring out how to minimize these jumps. I  would 
not be a big fan of contraptions that use two or three OCXO's to minimize  
the impact since all three could potentially jump at the same time. I would  
rather attack the problem at the source - most of the time its the  crystal.
 
There is probably something that can be done to minimize jumps, such as the  
crystal cut, modifying the crystal operating temperature, modifying the 
current  through the crystal, comparing different mounting methods, changing the  
direction (orientation) the crystal is mounted in, etc etc.
 
Here is a hint for experimentation: I once had an OCXO that jumped all the  
time (many times per hour). Since this was a reject part (definitely a factory  
RMA), I took a large screwdriver and gave the OCXO a good bang. The jumping 
was  immediately mitigated down to a couple of times per day. That causality 
was a  real eye opener, whatever stresses were on the crystal were released by 
the  shock.
 
bye,
Said 
 
 
In a message dated 10/23/2008 08:39:18 Pacific Daylight Time,  
quenbob5 at pacbell.net writes:

So why  would they not show the same kind of jumps as plain OCXO's?  Rick K.  
mentioned atomic standards avoid jumps.
Bob  Q.




More information about the Time-nuts_lists.febo.com mailing list