(This is an update for Temperature variations in TMP117 but since I had the wrong model of sensor, it needed a better title)
So I am actually looking at 3 BME280 sensors. I did some testing with all three sensors under a plastic tub with a couple of thermometers:
The temperature readings over around 2 hours were pretty consistent:
- Both thermometers ~ 18.1
- Sensor 1 ~ 17.6
- Sensor 2 ~ 17.5
- Sensor 3 ~ 16.5
The first question is why Sensor 3 is consistently lower than the other two. Any ideas?
The second question is why Sensor 1 and 2 are around 0.5 degrees below the thermometer. The BME280 has an accuracy: of ±0.5 °C so is that the explanation?
If sensor 3 is the one nearest the camera you will note there is a much larger gap underneath the plastic tub on this side. It could be outside air influence. To be accurate with the comparison all 3 sensors should be as close together as possible and as close to the thermometer bulbs as possible. The fact that the other 2 are a bit closer to the small gap under the other edge might account for a bit of the other 0.5ºC. Also the accuracy of the other 2 thermometers is unknown although they both read the same they seem to be the same make and model.
I would think that if you have everything within 0.5ºC you are doing pretty well. For home trials without a lab standard I don’t think you could expect much better.
Hey thanks for the reply! The one nearest the camera is sensor 2. The one in the top right is sensor 3, which just seems to be off. Not sure if I need to replace it, or try TMP117 sensors instead.
And yes, the thermometers are cheap ones from Bunnings, not sure of their accuracy.
Well that is not going to give any indication of absolute accuracy but they are at least both the same and close to your sensors.
But to get any meaningful comparisons the sensors have to be as close together as possible. As is you are not looking at any disastrous variation (1ºC) which you could get by breathing on one of them. I think any variation of a degree or so with devices outside a lab is not bad. If you want better I think could involve $$$. What are you going to measure anyway or is it just an experiment for interest.
Have a read of this thread:
Compare piicodev TMP117 and BME280 temperature sensors
Use the BME280 for pressure and humidity, do not use it for temperature measurement.
For accurate temperature measurement (+/-0.1degC) use the TMP117.
Hey thanks all, I switched to TMP117s and everything looks peachy. I’m keeping the BME280s for humidity, which they seem good at. Cheers
Thanks for updating us as it is useful information to have regarding any temp/atmospheric projects.
Best of luck with the project and would love to hear how it all turns out once completed.