[time-nuts] oscillators

[SAIDJACK at aol.com]

There are some drawbacks to this type of SC-cut MCXO I would think,  and it 
could possibly be replaced by a much higher performance Symmetricom CSAC,  
probably at a lower cost and higher availability:
 
* Q-tech MCXO is ITAR controlled, CSAC is not
* MCXO has 20ppb over temp stability, CSAC has 1ppb spec, typical units can 
 be much lower than that. Our GPSTCXO has 75ppb over temp, not much more  
than the MCXO, but probably at 1/10th the cost.
* The G-sensitivity of CSAC is at least 50x better than the MCXO  (its so 
low, its very hard to measure)
* Aging of MCXO is much higher (order of magnitude)
* Symmetricom is a big player, Q-tech is relatively small and unknown
* I remember that there are patents on the MCXO held by FEI?
* Power consumption is very similar, future CSAC units will have much lower 
 power than the MCXO.
 
The MCXO does have lower phase noise, it's an SC-cut cyrstal after all. But 
 with CSAC's now becoming available from multiple sources, why use an MCXO?
 
bye,
Said
 
 
 
 
In a message dated 8/30/2012 10:35:56 Pacific Daylight Time,  
richard at karlquist.com writes:

First,  it has to have no activity dips over the full
operating temperature  range.  OCXO crystals only have to work
at the oven temperature.   Activity dips, according to John Vig,
are the reason whey they don't use  mode B to sense temperature.

Second, it has to be cut correctly so that  the beat note between
the fundamental times 3 and 3rd overtone modes is a  useful
indicator of  temperature.