[time-nuts] Some More questions

[Chuck Harris]

Didier Juges wrote:

> One problem with that approach is that crystals that are not intended 
> for oven operation are optimized for minimum frequency change over 0-50 
> or some other "normal" environment temperature range, and at 75 degree C 
> or wherever you are going to run the oven at, the temperature 
> sensitivity might be much greater than around 25 degrees. So even though 
> the oven might reduce the temperature variation by a factor of 10 or 
> better, the overall frequency sensitivity may not improve by the same 
> factor..

Of course, one could take a small Peltier device, and a resistance heater,
and make a really good oven that keeps the crystal right at 25C.

By having the ability to both heat, and cool the oven, you can achieve
a much more precise control of the oven's temperature.  A conventional
oven has the ability to rapidly heat, but cooling is handled through the
leakage in the oven's thermal insulation.  The number of calories provided
to the heater, in a conventional oven, is precisely known, but the number
of calories sunk by the "cooler" (insulation leakage) varies with ambient
temperature.

-Chuck Harris