Temperature controller dependenced on lab ambient temperature

[David Ackerman]

To control DFB laser optical frequency, we use precision current controller and temperature controller. The question here is: what is the dependence of the controller's temperature set-point on lab ambient? In other words, if the temperature of the temperature controller changes by 1 degree over the course of a day, how much will the set-point, and hence the device under test, change in temperature (without showing any change in error signal at the controller)? Of course, we are using a particular vendor's controller, but I'm interested in general. If your answer depends on the particular controller, I'll be more specific.

-- Thanks, DAA

[Steve Gwinner]

It would be very difficult to accurately specify temperature coefficient for a thermoelectric temperature controller. While precision current sources have temperature coefficient specifications on the order of <75 ppm / ºC, the response is linear so specifying this is straightforward. But for a TEcontroller, a change of +1 ºC in ambient may also influence the temperature sensor at or integrated into the DFB - which will force the controller to increase power to the cooler. After the device is cooled the TEcontroller will theoretically settle back to it’s previous power output level. There would need to be some time function in this calculation, and the response would not be linear. Also, because the feedback from the temperature sensor determines the response of the TEcontroller - you would need to consider more closely the temperature coeffecients associated directly with the the sensor. You would need to consider the temp coeffecient associated with the source current through the thermistor, and the voltage measurement (the instrument sees) for the feedback. You would expect a change in thermistor resistance for a 1°C change in environmental temperature. If there is a temperature coefficient associated with either or both the thermistor current and voltage measurement which would change the value of either, then I would expect a change in the set-point temperature depending on the magnitude of the temperature coefficient. Having said all of this, I believe that the affect of any temperature coefficient for thermoelectic temperature controllers (ºC / ºC) for the sort of changes in ambient you have mentioned should be negligible in terms of impact to your application.

[Larry Johnson]

I agree with Steve that it is difficult to specify a temperature controller's sensitivity to changes in ambient room temperature. At ILX we use two levels of temperature control for obtaining highest stability for DFB lasers. We control both the case temperature and the internal TEC temperature.

Whatever you do, probably the simplest approach to understanding the sensitivity to ambient temperature fluctuations is to just measure it by adjusting the room thermostat up 5C and recording the resulting change in your DFB laser output.