How to keep sensors cool in hot weather?

Thread Starter


Joined Jan 2, 2017

I am using rather temperature-sensitive NO2 gas sensors for air pollution measurements. Thing is, I am in Houston, TX, and it's pretty darn hot down here. Collecting data over a period of 24 hours showed that at high temperatures (more than 32°C or 90°F) the sensor's accuracy plummets, the readings also get significantly offset, and all of that is non-linear. Calibration is pretty tough, and I'd rather just cool it down.

The gas sensor comes with a noise-reducing board; therefore two things might be suffering from the heat: the sensor itself and the noise board. Albeit the ICs on the board don't seem to heat up too much.

Hence I am trying to figure out how to cool both of these devices (and the surrounding air) down. I am looking for a portable solution.
Fans and heatsinks are of no use - the ambient temperature is already too hot. I am in the process of trying thermoelectric coolers - peltier; though they seem to be useful only when in contact with a chip, which in my case is impossible; is it even feasible to cool down a small volume of air with peltier coolers?

I could also put the sensor in a small chamber and then keep the air cool inside (thermally isolating box + peltier), and then pump external air into the chamber.

I am just kind of stuck on this one; I would really appreciate any insights, suggestions, and advice:) Are peltier coolers appropriate in my case? Any suggestions for thermally isolating tapes/materials for a chamber? Is there a better way to solve this one in the first place?


// datasheet for the sensors:
// and for noise reducing board:


Joined Jan 15, 2015
A Google of 12 Volt Coolers will bring up dozens of the things on Amazon but they are for the most part large and bulky where you just need something small likely no larger than a soda can. You may be able to use a few small Thermoelectric Cooler Peltier type devices designed for low voltage DC but the problem becomes they have high current demand. This is a problem when you rely on battery power. Hopefully another member has been there and done that. You really only need a small chamber to place things in. You could possibly experiment a little with a few inexpensive TEC devices, some insulating foam and small containers.



Joined Mar 19, 2019
A quick scan of sensors shows yours is about average temperature wise. There are some higher rated models. For example: rated for 55°C

Didn't find the PDF but general specs are a bit higher temp than others. A better-rated sensor may be your answer. Even yours is not that tight as temp goes up. Just what are you measuring? Ambient air? Flue gas? Portable grab samples?


Joined Apr 11, 2010
You mention that thermoelectric coolers - peltier modules - are only effective when in contact with a chip.

Attach the module to an aluminum block or even a heat sink. Mount your sensors in an insulated space and insert the aluminum inside. The peltier cools the aluminum and the aluminum cools the inside of the box.


Joined Mar 19, 2019
I have also seen where the instrument tech pinched down a piece of copper tubing to make an air jet supplied by compressed dry instrument air to aim at the device for cooling down here in coastal Georgia. Even wet compressed air works if moisture not a problem. Even better if a heatsink also used.