Temperature-following adiabatic shield

I've constructed a custom calorimeter for research purposes. In order to increase the accuracy of temperature measurements, I need to build what is called an adiabatic shield. Funny name, simple idea. If the temperature of the calorimeter contents are the same as the surroundings, then no heat will flow in or out of those surrounding materials. Thus, a measure of the internal heat energy may be made to higher precision.

The job of the shield is to follow the temperature of the contents in the calorimeter. The calorimeter will operate at room temperature. We're talking about a maximum heat rise of 10-15 C over 2-10 minutes at the very most. A delta T of + 2 to 4 C is more likely. Manual temperature following is possible though tedious and the accuracy is likely to be poor with a good possibility of excessive temperature lag or overshoot. I've scoured heaven and earth for circuit designs that can replace the few vacuum tube versions I found from NIST.

The intended surface to be warmed is a 4 inch diameter circle made of brass.

An Arduino Uno board is already being used for other elements of operation, but I'd really like to do the temperature following with another Arduino board controlling a resistive element such as a ~10W power resistor.

Question: Is there a resource out there I can access for such a temperature following circuit? I could use a thermistor, PT100, thermopile or if absolutely necessary a thermocouple sensor. Any good ideas out there??

Thanks in advance!

Ansa Metallikker

The type of temperature sensor will depend on the required accuracy of the measurement, and the size of sample compartment, which you don't state.

To have the temperature of the brass plate follow the temperature of the sample compartment, you will likely need to measure temperatures at several different locations.

Look up the old national semiconductor linear applications databook - there's a nice examle there.

It's online


Thanks for responding. The brass plate covers the opening of a 600 mL cylindrical Dewar with a 2.75 inch inside diameter. The primary temperature sensor inside the calorimeter is tentatively an independently operated Pt100 sensor. The Pt100 has quite good sensitivity and precision. But perhaps a thermistor is a better choice for both temperature sensors.

My sense is that a matched thermistor on the adiabatic shield (R2) would be the optimum choice for precision and low thermal mass. This project is a series of successive approximations and improvements. If the adiabatic shield could be within 0.5 C at equilibrium, that would be a good start.

The temperature-following circuit could use the resistance of both thermistors for some kind of null detection as in a Wheatstone bridge. As long as the primary sensor (R1) is left unchanged. The question is, how to get the Arduino to bring the heater along to null out the R1-R2.

Thanks for whatever guidance you may be able to offer.