BME280 breakout boards all showing humidity 60% too high

I2C, SPI, ESP32, Arduino Mini Pro, Pi PICO, it doesn’t matter how I connect it or what I connect it to. I bought 5 of them and they all read 60% high for humidity. I have tried Sparkfun’s libraries, Adafruit’s libraries, same readings. I have two different (4-pin and 6-pin) BME280 boards from ebay and they all work as expected (correctly).

Any ideas?

Are they being mounted/used differently?

Are the ebay ones using real bosch sensors?

How do you know they are reading high?

They are all being mounted on a breadboard. I live in Tucson (very dry climate), the humidity readings of the Sparkfun boards are reading above 90% humidity. DHT22s and other BME280s are all reading below 30% and are in agreement with each other within a few percentage points.

Share a photo of the wiring/setup

Interesting. You think the breakout board would read temperature and pressure accurately but not humidity and it would be because of the wiring?
I’ll snag a photo tonight and get it up here. Do you want a photo of I2C wiring or SPI? Which board do you want it connected to?

There’s a fair amount of burn-in time on those, and they are sensitive to how air circulation is around 'em/mounting/environment considerations

I’m asking for a photo to try and see any differences in those, mostly. Wiring doesn’t matter much in this case (I just ask for both by default)

Are the other sensors being used in the same location?

Do the readings drift one way or another over time?

They are being used in the same corner of the same room, even on the same breadboard. I have not monitored them to see if they drift in a direction, but I can do that. Originally, I had them recording readings every five minutes to a google sheet. I’ll turn that on again tonight so we have some data to work with.

I found the spreadsheet. Over a period of 2.5 hours, the readings did not trend in either direction. I’ll still grab a picture of the wiring and start the readings again so we have some current data.

The sensor has been running for three hours and did not move the first two hours. It dropped 2.5% over the third hour, but that is not uncommon in our living room.

Pullup resistor makes no difference.

Maybe a stupid question but is the Promini VCC 3.5V or 5V?

A number of BME280 boards accept 5V, but Sparkfun expects 3.5V

This pro mini is a 3.3v board. I have only connected these breakout boards to 3.3v sources. Good catch though, I appreciate your help!

Maybe they are all damaged, maybe during transport. Another “out side the box” thought, because they all fail : could it be that BMP280 have been populated instead of BME280? The BMP280 does not measure humidity BUT is used with the same library.

After successful calling mySensor.beginI2c(), try calling

uint8_t chipID = mySensor.readRegister(BME280_CHIP_ID_REG); // Should return 0x60 or 0x58

0x60 = BME280 0x58 = BMP280

Yeah, I thought it was weird that they were all wrong as well. The cheap ones I got off ebay all work. So, I thought it was weird that the ones from Sparkfun that cost 5 times as much were not working as expected. I have decided to just use the ebay sensors for my project, and they are working well. In order to get a refund from Sparkfun, I had to post the issue here…

Oh, I do get 0x60, so I think it is a BME280

yep.. it is BME280… it also means that the I2C communication is working.

I am coming back on the 5V to 3v3. Just to make sure : did you measure the VCC ? The only difference between Pro mini a 5V and 3v3 seems to be the clock speed (8 of 16 Mhz) . Your FTDI should be set to deliver 3v3

Else

Unless there is some sort of sealing on top off the BME280 and you have done the long burn in… I suspect the BME280 on the boards are damaged.

The pro mini is the third board I have connected these to and they all give the same results. Originally, I had them connected to a raspberry pi, then I connected them to an ESP32 and finally the pro mini, just to see if it made any difference. All boards report the same odd humidity readings. I wired them SPI as well and received the same results.

did not know that you had done all that testing and gone through that hassle … looks to me the BME280 boards are damaged.

Sparkfun has to step in…

Thanks for the info; it sounds like you’ve filed for an RMA already but in case you haven’t:

Was it purchased directly from us? If so head over to Return Policy - SparkFun Electronics (contact vendor if purchased elsewhere) and we’ll get ya squared away

Thanks, Homie. I appreciate all your help!