[time-nuts] Re: Long term stable environmental sensors

[Dan Kemppainen]

Attila,

Has this board been restarted lately? Does the temperature error exist 
after a reboot of the hardware/software?

One very long shot/minor concern is errors in the calculation in the 
micro due to bit flips or other silliness with the calibration 
coefficients in the micro controller memory. (This would only be a 
concern for devices where the coefficients are only read at startup).

Is it possible there is actually more heat near the sensor? New 
computer/phone charger, or failing power supply near by?

I'm also curious how much these drift, as I've been running them in the 
lab here too.


Dan



On 11/29/2022 10:22 AM, time-nuts-request at lists.febo.com wrote:
> Moin,
> 
> I have been monitoring my lab with a homebrew sensor board based on a BME280
> combined temp/hum/pressure sensor from Bosch. Unfortunately, the sensor seems
> to have quite some drift. Now, 4 years later, one of them shows 30°C when
> every other thermometer in the room shows ~22°C. And as temperature is
> needed to for compensation of the other measurements, I can't trust those
> either.
> 
> So, now I'm about to design and build a new system that has better long
> terms stability.
> 
> At the moment I think I will go with separate sensors for temp, humidity
> and pressure. Temperature is "easy", a PT100 will do the job nicely.
> For humidity, there is the SHT line from Sensirion that is well known.
> But I don't know of any decent absolute pressure sensor that is long-term
> stable and doesn't cost an arm and a leg.
> 
> Does someone know a pressure sensor that is long-term stable?