[time-nuts] Another "atomic" clock question

Hal Murray hmurray at megapathdsl.net
Sun Mar 2 12:56:01 EST 2014

bob91343 at yahoo.com said:
> But I thought conventional wisdom is that most crystals are AT cut and an
> attempt at zero average coefficient is made, causing a nonlinear
> characteristic.  But perhaps over a limited range it's linear.  The problem
> of course is calibration.

Most crystals are low cost.  They will have a temperature characteristic 
similar to the graph about half way down this URL:
The specs on the standard oscillator packages vary from 100ppm to 20ppm.  
That covers temperature and voltage and initial manufacturing and some amount 
of aging.  (I haven't looked at the spec sheets recently.  I don't remember 
seeing anything about aging.)  The point is that they are low cost and the 
specs are reasonably clear, something a digital designer can understand and 

> Again, how does one calibrate those 3 MHz ovenized units?

I plug mine into a HP 5334B which is clocked by a TBolt.

These are my opinions.  I hate spam.

More information about the time-nuts mailing list