I am building a device to heat and measure the temperature of a small vessel (double boiler) of gasoline over a range of temperatures say -10F to 120F. I will be measuring the capacitance across a homemade probe to figure out the dielectric constant over this temperature range. What would be the best way to accurately log the temperature while simultaneously measuring the capacitance. Currently, I will be using a 555 timer and microcontroller to measure the capacitance so the same microcontroller could take an analog input from a temperature probe. Would a thermistor work best in this situation? Can I put a thermistor directly into gasoline safely? Would it be better to consider a thermal imaging approach?

  • $\begingroup$ I will be conducting the experiment outside and will place the assembly in contact with a large mass of scrape steel $\endgroup$
    – piman
    Commented Mar 22, 2017 at 17:11
  • $\begingroup$ Two things to add: Make SURE you have taken proper grounding into consideration for your apparatus; gas + spark = no eyebrows (or worse!). Also, make sure your thermistor is suitable for that range of temperatures; the rule of thumb is to double check temp. range on thermistor because accuracy is a strong function of the operating range. $\endgroup$
    – J. Ari
    Commented Mar 22, 2017 at 19:12
  • 1
    $\begingroup$ Heating gasoline. WCPGW? $\endgroup$ Commented Mar 22, 2017 at 19:20

2 Answers 2


A thermistor will work fine. Most are epoxy dipped, and your test is for a relatively short period of time. A theromcouple would work well also, but they take more supporting electronics for signal amplification. As long as your wires running to your thermistor are not to long, it will work great for this temperature range.
10k NTC thermistor with lug
10k NTC thermistor with 1% tolerance

A data acquisition unit is typically the way engineers do data measurement since using a mictrocontroller requires extra work. If you have it already set up because you will be using this capacitance sensor on said microcontroller in the future that would be a good way to go.

If you go the data acquisition (DAQ) route, I recommend Labjack as it has drivers for opensource python. You would feed both the capacitance (probably a circuit required or even output from your microcontroller), and your temperature to the DAQ unit. National Instruments makes good DAQ units too, but plan on running those with LabView.

One of the temperature loggers Gwydionforge recommended in his answer would work, but it is very difficult to correlate data captured on two different systems. The data will have to be combined in a spread sheet or scripting program. If the time stamps are not identical you will have to resample (interpolate) to get a capacitance and temperature at each timestamp.

Soapbox safety speech: Gasoline vapors are very flammable; do your test outside or in a hood. Make sure to wear safety glasses and a face shield.


I would use a thermocouple in the gasoline to a data-logger. You can get simple data-logger from omega (Temperature & Humidity Data Loggers)

It will be safe to put in the gasoline, just be sure to use a good fume extraction hood that is suited for combustable vapours and keep all metal in contact with the gas grounded and bonded to prevent sparks.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.