[time-nuts] How to accurately measure an oscillator's temperature.

David J Taylor david-taylor at blueyonder.co.uk
Thu Apr 24 00:48:09 EDT 2014

I have both an OCXO and an FE-5680 Rb oscillator and I'd like to track
their temperatures.

What is the best why to measure?   Maybe each has a different best method

The OCXO is just a small steel can.  Is measuring the steel can temperature
the best why to go.  Epoxy some kind of sensor to it?

The Rb is mounded to a large heat sink and there is a fan.  I want to
control the fan so as to keep the Rb temperature constant.

In both cases I tried using TMP36 three terminal sensors and just got
noise.  The reported temperature was up and down more than 2C.    The fan
controller just chases noise.

BTW the fan based temperature control is effective.  The FE5680 gets very
warm in it's box but if I give the 12V fan even 8 volts the heat sink
quickly cools.  I want to throttle the fan to keep the Rb at constant
temperature but the temperature data I'm getting is not very good.

The problem I think is that any sensor I have is on the outside of the
oscillator and is effected by cooling air   What are others doing?   What's
the best kind of sensor.

Chris Albertson


You could try the DS18B20 "single-wire" device.  Maybe it's better, maybe 
not.  I've played with it on a Raspberry Pi:


You can set the precision to 1/16 °C, although the accuracy is only 2 °C. 
You should not be seeing just "noise", though.

SatSignal Software - Quality software written to your requirements
Web: http://www.satsignal.eu
Email: david-taylor at blueyonder.co.uk 

More information about the time-nuts mailing list