Self-heating thermistors for flow measurement


Thread Starter


I am looking for any information on the circuitry involved in operating a thermistor as a flowmeter. I am aware of two different techniques:

i) keeping the thermistor at a constant temperature and measuring the power needed to maintain that temperature

ii) keeping the power constant and measuring the heat loss from the thermistor

I believe the first circuit is a fairly complex (expensive) control circuit and I would be more intersted in the latter option. If you could
give me any more information on this application I would greatly appreciate it.


Bill Gausman

There is another method called a "Constant Differential Thermal Anemometer" that uses two thermisters. The first is used to measure the airstream temperature, the second is self-heated to a setpoint some fixed value (perhaps 10=B0) above the airstream temperature. An example is the SM2-SVP sensor manufactured by Staefa Control System. The other methods have errors introduced as the airstream temperature changes. This is an inexpensive and highly accurate, albeit nonlinear, method of measuring air velocity, especially at the low end.

Bill Gausman, PE
System One Control, St. Paul, MN
I would also like some circuit ideas. I remember years ago a National Semi app note, but I cannot seem to find it. I guess I will try to pulse a 10K NTC with a MCU and read back dr/dt with ADC, and then calibrate with known flows.

Cullen M. Sabin

The Penn State University did quite a lot of work on thermistor anemometers back during the early 60's. I was involved with hot wire anemometry (so-called) for operation in water at Stanford in order to do some research. It was supposed that I could learn something from them so I made a trip to visit. As it turned out I was able to help them.I built and sold a few hot wire instruments for water research in the 60's, and also built a few hot thermistor yacht speed indicators. The feed back circuits for operation at constant temperature are really very simple. I used analog operational amplifiers. The time constant for thermistors if fairly long. The guys at Penn State were making their own, about the size of a pinhead, but they were still too slow for good air work.

Cullen M. Sabin
[email protected]