Just received my Kit to build a Enviro Weather station and it has prompted me to ask about an issue I experienced with my Enviro+. If the sensors were directly connected to the PI Zero the heat from the Zero effected the readings. I had to connect them remotely to stabilise them.
Does the Enviro Weather suffer this as it is back to back with the Pico W ?
Thanks
Hi, I’ve the PEW since a few months and I definitely see the influence of the PCB temperature on the BME280 readings. In the programs (config_defaults.py ) there is a provision for taking care of that, when the board is powered by USB and not by the batteries. Unfortunately, in my case, the correction factor DEFAULT_USB_POWER_TEMPERATURE_OFFSET = 4.5 is not enough to compensate the temperature measurement drift even when I use a LiPo battery to power the device. I’ve tested that building a PEW like thing with only the BME280, the anemometer and the wind vane. Note that the BME280 is connected by wires to the Pi Pico and it floats in the air, a few centimeter from the MPU.The temperature reading exceeds by 1.5 - 1.7 Deg C what is measured by a DHT22, a DS18B20 and an Oregon Scientific Hygrometer - Termometer remote unit… In addition also the humidity measurement differs from what BME280 and DHT22 measures in free air a few centimeters from the Pi Pico board.
I would also check the pressure reading as , there is the usual sea level compensating factor but, again, it seem not to be effective (10 - 15 hPa less that what is reported by an official meteo site located in my town). I’ve developed the impression that the BME280 has got a lot of hype but it needs to be carefully compensated for temperature to get reasonable reading for humid and pressure. Ah, by the way, in the BME data sheet is clearly written that the sensor is not meant for accurate temperature measurement and that temperature readings are need for compensating pressure and humidity ones. In a near future I’m going to post request for help on the blog to see if anybody has solved this lack of precision or if I’ve no luck with this PEW and the sensor is defective.
P.S. the data I’ve collected seems to show a non linear dependence of the BME280 readings from temperature (please note that my setup could be approximate or even misleading). This would mean to me that you cannot simply use a number to correct the influence of the PCB temperature but you need a curve from which the program should take the right correction factor.
Interesting post Fabri54.
As the BME280 is onboard and not separate we can’t remove it from the board with a ribbon cable as I did with my Enviro+. I think that only leaves one solution:
Careful calibration and creating a correcting equation. Could be very time consuming to achieve as the equation would probably have to include ambient temperature. The board might also need to be gently fanned to even its temperature environment. Sadly all very complicated.
I have not been able to assemble my Weather Station so can’t provide any practical input into this, perhaps some other users can?
Many thanks for the reply. Frankly speaking, I’ve reached the conclusion that the PEW is just a little more than a toy. Good for learning the ropes of a meteo-station (FW and HW) but nothing else. In addition the anemometer and the wind-vane are quite lite, made in a plastic that may not last long under the sun (mediterranean sun).
I’m afraid that the idea of fanning the PEW may not be a viable one. In principle, a barometric sensor should be placed in the shadow and in still air. I’m mounting the PEW in a Stevenson Screen that, somehow, meets these requirements. What I don’t like is that the humidity inside it might not be correct, lower, as the PCB is heated and that might lower the RH%
This Spring I believe I’ll mount it, mostly out of curiosity, and see how it behaves in presence of some strong winds. The Mistral can normally blow at 60-70 Km/Hrs but it can peek at 100Km/Hrs. We’ll see
Yes, I think you are right, not much more than a toy, but it provides a glimpse into weather measurements and how carefully they should be made. I think the Enviro+ might have a better chance of accuracy. I have spent some time with that and had an interesting discussion in this thread.
https://forums.pimoroni.com/t/envroplusweb/12930/152
Like you, Chris has Atlantic/Mediterranean type weather to deal with.