[ntp:questions] how do I lock in average frequency correction

Terje Mathisen "terje.mathisen at tmsw.no" at ntp.org
Mon Feb 13 07:31:27 UTC 2012

Richard B. Gilbert wrote:
> Rather than cooling the crystal it's customary to to put the crystal in
> an "oven". This is not the oven usually used for cooking. What it does
> is to heat the crystal to a temperature that can be maintained 24x7
> and will be a little warmer than the highest temperature that can be
> expected naturally. A value in the range 120 to 130 degrees F could be
> used.

The 10 Mhz OCXO which is the frequency reference for a cell phone base 
station I'm working with is probably keeping the oven at a slightly 
higher temperature:

The base station itself is specified to tolerate ambient temperatures up 
to 55C (135F), so the oven has to keep the crystal a bit higher, 
probably in the 60-70C range?

Anyway, when I googled OCXO the first link I found specified such 
crystals with shorterm (day) drift rates of 0.1 ppb, and a yearly drift 
rate of maximum 20 ppb.

That kind of stability would make a pretty good ntpd baseline as well.:-)


- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"

More information about the questions mailing list