💬 Atmospheric Pressure Sensor

[mfalkvidd]

@grumpazoid If you don't use the forecast feature, remove #define GENERATE_FORECAST. Doing that should give you much less ram usage.

[grumpazoid]

@mfalkvidd Thanks for that. This is going to be my outside weather node running off a 12V solar charged battery. I started wi a DHT22 but it failed after a few months.

[grumpazoid]

@mfalkvidd I've got it up and running now on the bench complete with measuring voltage. I found that removing all the serial print statements also free up a lot of memory. Thanks to all those involved for the provided code.

[Iliah Borg]

When using BME280 breakout https://lowpowerlab.com/shop/product/185 with Embedded Adventures library, life is easier if addrBME280 is set to 0x77 (it is set to 0x76 in BME280_MOD-1022.h)

[Fear na Boinne]

Weird question, but has anyone dealt with thermal isolation of the BMPx8x modules?

To increase the WAF, I have one, together with a BH1750 and an rf-nano in a small, aesthetically pleasing, 3D printed custom housing. However, the temperature is stuck at 26.2°C, which I assume is due to the radiant heat from (mostly) the Arduino and the nRF24L01+

So I am looking for ways to keep the BMPx8x cooler, so it can correctly measure the temperature (and therefore also the pressure, as the warmer chip means the barometric adjustment will be off)...

[mfalkvidd]

@Fear-na-Boinne how about calling sleep() between measurements? That will pus the Arduino and the nrf24 in sleep mode, where they won't generate heat.

[Fear na Boinne]

@mfalkvidd Didn't think of that... I can try changing the delay() to a sleep() and see how much that impacts the temperature...

(But I still am open to ideas to help insulate the sensor, and am now also wondering whether creating extra vent holes on both sides of the rf-nano might be a good idea, even if it means more dust can come in as well...)

[BearWithBeard]

@Fear-na-Boinne I don't know how your device / enclosure is built, but it's good practice to place environmental sensors as close to an opening in the enclosure and keep the dead volume around it as small as possible. Ideally, you'd put it in a separate chamber and move it as far away from other electronics and possible heat sources to minimize their influence to the measurements. If you're worried about dust, you could use thin cloth or something like that around the openings as a dust filter.

That being said, the BME/P sensors tend to report too high temperatures (just google for "BME temperature too high" or something like that - you'll find tons of topics), partly due to self-heating depending on how you use it, e.g. with oversampling enabled. IIRC, using forced mode, oversampling at 0 for pressure and 1 for temperature (and humidity) and filter turned off, is the prefered way for simple weather measurements with minimal current draw according to the datasheet.

If you have one of those 5V modules, the voltage regulator on board might give off some heat aswell.

The "easiest" fix would be to offset the temperature reading in software by a few degrees, if the reported temperature is off by the same value over the whole relevant temperature range (assuming 15 - 30°C for room temperature in your case).

[Fear na Boinne]

The min temperature has dropped significantly since using sleep(), but still the lowest I have seen is 22.8°C, when the other sensor in the room (The central heating thermostat that speaks OpenTherm) was showing a bigger variance.

My sensors are running on the 3.3v the Arduino puts out, so no extra regulators putting out heat (ie other than the one on the Arduino board).

I designed openings and stand-offs in the lid of the enclosure for the sensors, and the rf-nano sits at the bottom of the enclosure, but atm they share the same single chamber... Thinking of drilling a few venting holes near the rf-nano to let warm air out and cool air in...

IIRC am not oversampling, just doing timed single samples, but I'd have to check that.
It's currently sampling once every 10s because the enclosure also has a BH1750 lux sensor I want to sample more frequently, and they obviously run in the same sketch... :P

[Fear na Boinne]

Putting the graphs next to eachother, the graphs looked similar, and after guesstimating the difference and applying that to the sensor as a correction in Domoticz, the graphs are overlapping sufficiently to explain the differences by the difference in location (one is in the center at ~1.7m height, the other on a shelf in a corner at ~2m height) and standard deviation...
I'll keep it like this for a while and will reevaluate at a week's interval.

FWIW I am actually doing the basic oversampling, while doing the timed single samples.
If it remains an issue, I could modify the code to start the sensor with reduced oversampling/accuracy setting, but for now it looks good (enough)...