[time-nuts] Low-long-term-drift clock for board level integration?

Tom Van Baak tvb at LeapSecond.com
Mon Feb 20 06:38:40 UTC 2012


> Here is a typical high end OCXO.  (It may blow your budget, but we can use it 
> as an example.)
>  http://www.mti-milliren.com/ocxo_270_ocxo.html
>  The typical 5 MHz aging performance is 5E-10
>  per day and 5E-08 per year.
> 
> That's not in the right ballpark.

Here's another way to look at it.

A typical quartz frequency aging rate of 5E-10 per day means you
could be off by 5e-10 in frequency a day later, which means you
are now off by 20 microseconds in time at the end of the first day.

Time error grows linearly with frequency error and quadratically
with linear frequency drift (time error = fe * t + 1/2 * fd * t^2).

After 2 days you will be off by only 1e-9 in frequency but a total
of 80 microseconds in time.

Assuming the linear frequency drift trend continued, at the end of
a week you're under 4e-9 in frequency error but the time error will
now have grown to 1 ms! And this is under laboratory conditions...

So this is why you have to use an external timing reference.

/tvb





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