I have an HIH6130 temperature and humidity sensor inside a plastic enclosure with a Particle Photon.

When it takes reading while inside the enclosure, the readings are ~ 2 degrees too high, vs outside the enclosure.

Outside of putting a fan or a heat sink on it, any other ideas on how to reduce the heat buildup?

  • \$\begingroup\$ If you're putting a fan on it, you might as well just lose the enclosure. \$\endgroup\$ – Marcus Müller Mar 14 '18 at 1:34
  • 2
    \$\begingroup\$ I like the heat sync idea. It syncs the temperature around it with what it is trying to cool down. \$\endgroup\$ – Harry Svensson Mar 14 '18 at 2:03
  • 2
    \$\begingroup\$ Usually sensors are located at the bottom of a vented enclosure or protruding so that convection get external ambient flow and less influence from self heating. It ought to be the THT part and not SMT to locate the part in a better location or simply calibrate an offset \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Mar 14 '18 at 3:23
  • 1
    \$\begingroup\$ @DavidSchwartz making the experiment fit the data or the data fit the experiment : called Finagle’s constant.. :) :) \$\endgroup\$ – Solar Mike Mar 14 '18 at 7:00
  • 1
    \$\begingroup\$ @DavidSchwartz, the correction curve might work, but to a degree. The result, unfortunately, will depend on geometry of the enclosure, and its orientation with regard to the field of gravity. Which likely will be difficult to enforce in appliaction field. \$\endgroup\$ – Ale..chenski Mar 14 '18 at 7:18

The "Particle Photon" board (who makes these goofy names?) consumes 80 mA at 5V, making its power dissipation at 400 mW. This is a pretty substantial power consumption, and the heat must dissipate via build-up of temperature difference between the board and ambient air. If you want your sensor and MCU inside one enclosure, you should find a way to reduce the consumed power.

One way to reduce the power of a MCU is to use aggressive power management, to put the MCU into sleep mode as soon as you can afford, and not wake it up as long as you can. The time sampling for temperature and humidity is likely in several seconds (given thermal mass of typical PCB and enclosure), so there must be plenty of opportunities to exercise deep power saving modes.

| improve this answer | |
  • \$\begingroup\$ Thanks, that makes a lot of sense. I'll try the software solution first as the readings are taken once every 5 minutes, so that's plenty of power reduction. \$\endgroup\$ – Misha M Mar 15 '18 at 17:47
  • \$\begingroup\$ @MishaM, good, just make sure to take data readings first while the board is at ambient, before running any radio communication stack. \$\endgroup\$ – Ale..chenski Mar 15 '18 at 18:43
  • \$\begingroup\$ Right, that makes sense. I'm going to try to put the antenna on the outside and disable the chip antenna to see if that helps as well. \$\endgroup\$ – Misha M Mar 15 '18 at 19:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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