conductive glue for BME280

[cmumford]

I'm putting together a little board with a BME280 temp/humidity sensor. It seems that the humidity readings on the various BME280 breakout boards that I've purchased don't accurately measure humidity. I've never seen a humidity < 40% or greater than 60%. Reading the datasheet it looks like if the BME280 gets hotter than 60C (which it will during soldering) then "reconditioning is recommended". Am I reading this correctly? I can't create a low humidity environment to "dry bake", so I'm wondering if using conductive adhesive would work and allow me to avoid soldering that one component. Thoughts?

[Ian.M]

according to https://www.bosch-sensortec.com/media/boschsensortec/downloads/handling_soldering_mounting_instructions/bst-bme280-hs006.pdf
the reconditioning procedure is:

Why would you think that's difficult? Start with ordinary room temperature air at a normally acceptable humidity, and heat it to 120 ºC and its R.H. will certainly be under 5%. so any temperature controlled electric oven will do, or even a small cardboard box on an electric hotplate.

Getting 75% humidity at room temperature is also easy - simply wet Sodium Chloride (common salt) with a little water and mound it up in a cone in a shallow dish so both the crystals and the liquid surface are in contact with the air.  Put the dish of salt in a sealed container together with the sensor(s) you need to recondition.

ref: https://www.engineeringtoolbox.com/salt-humidity-d_1887.html
and: https://www.conservationphysics.org/satslt/satsol.html

[cmumford]

I was mostly worried about being unable to get to 75%. If it said >75% then I'd feel more comfortable about the re-hydration step. Thanks for your suggestion though. I'm building multiple boards so I think I'll do an experiment and compare different approaches.

[Ian.M]

Note  that it takes significant time for the salt/salt solution's equilibrium with the R.H. of the air to stabilise.  That means you need to keep the sealed container at a near constant temperature, so in the absence of a temperature controlled environmental chamber, put the sealed container inside a styrofoam cooler, and fill spare space in the cooler with plastic bottles filled with water that have been kept in the same room for  a day, as a thermal buffer.

It would certainly be worth monitoring the temperature and humidity in the sealed container so you can see how stable they are.

[floobydust]

BME280 comes with a kapton hole cover to prevent contamination from soldering fumes and pcb wash. Don't take it off before.
The older (Bosch) software libraries had a bug where the BME280 calibration constants were not properly loaded on power up. Then everyone copied these libraries as their own and quite a mess. So I noticed simply playing around changing libraries I found also changed the readings.
As I recall the fix was adding a 10msec delay, but I still suspect something else is at play. The temperature sensor is crude, just a diode and I find it's way out. Although some of the boards I used are from Aliexpress and I have not tried swapping in a genuine BME280 yet to see if that is the issue, that they skip the calibration of them. Barometric pressure is very stable.
Also I read the part can be sensitive to light.