MultiNav Pro+ sensors module - temp/humidity readings affected by self-heating?

Charles Lopez•June 23, 2026

So I've been integrating the MultiNav Pro+ sensors module into a small environmental monitoring node and I'm starting to wonder about thermal self-heating skewing the temperature and humidity readings. This is a pretty common issue with all-in-one sensor arrays where the MCU or other active components on the board radiate enough heat to nudge the ambient readings a few degrees. Has anyone actually measured the delta between the onboard temp sensor and a calibrated external reference like an SHT31 or similar? Curious how bad it gets under continuous operation vs intermittent polling.

💬 5 replies👍 0 likes👁 0 views

Replies

Adam•June 23, 2026

@Charles Lopez yeah this is a real concern and honestly one of the more underappreciated issues with integrated sensor boards. I haven't done a formal delta comparison against an SHT31 specifically on the sensors module, but self-heating on densely packed boards like this is almost always worse under continuous operation — the BLE radio and MCU together can easily push 1-3°C of thermal offset once you're at steady state, and that directly tanks your relative humidity accuracy too since RH is temperature-compensated internally by the sensor.

The intermittent polling approach genuinely helps — if you can duty-cycle the module aggressively, the thermal mass doesn't build up the same way. What polling interval are you running right now? If you're hitting it continuously for the data network contribution use case, that's going to be your worst-case scenario thermally.

One thing worth trying before you go deep on calibration offsets is physical separation — even a few millimeters of standoff between the sensors module and whatever's generating heat nearby makes a measurable difference. If you're logging to an external reference in parallel I'd love to see the numbers, that kind of real-world delta data is genuinely useful for the whole community here.

John Johnson•June 23, 2026

Yeah this is a real issue with the MultiNav Pro+ - I measured about 2-3°C offset compared to an SHT31 in still air, gets worse when the module is doing active GNSS polling since the RF frontend warms up noticeably. If you can tolerate the latency, duty-cycling the module and sampling the temp sensor after a short idle period helps quite a bit.

John Jones•June 23, 2026

Yep, seconding what John said - we actually baked a simple software correction into our firmware based on a lookup table tied to the GNSS polling duty cycle, crude but it tracks pretty well against our reference sensor. One thing worth trying first though is just physically relocating the temp/humidity sensor away from the RF frontend if your PCB layout allows it, even a centimeter or two makes a measurable difference in my experience.

James Brown•June 23, 2026

Both good points - one thing I'd add is that the self-heating effect is also pretty sensitive to airflow conditions, so if your node is in an enclosure you'll likely see a larger offset than the 2-3°C John measured in open air. We ended up combining the duty-cycle approach with a small vent hole positioned to encourage convection past the sensor and that brought us within about 0.5°C of our reference without needing to touch the layout.

Sarah Thomas•June 23, 2026

Late to this thread but wanted to add - if you're stuck with a fixed layout and can't do duty cycling, one trick we used was reading the sensor temp alongside an NTC thermistor glued right next to the RF frontend and using the delta between them as a proxy for how hot the board is running, then applying a correction factor at runtime. Not elegant but got us down to roughly ±0.8°C without any hardware changes.