Barometric sensors for volt-nut logging

[Cerebus]

A few weeks back branadic got some anomalous readings on a drift measurement that he was making on an LM399 which he put down to the foul weather we were having at the time, with winds so strong that the barometer was bouncing up and down. It just happened that a few days beforehand I'd realized my notepad (a Nexus 9) has a barometric sensor and had installed an app that sits in the background and keeps a rolling chart. That provided some useful evidence at the time that branadic's suspicions were probably right.

Prompted by that, I investigated what was available in the way of easily obtainable MEMs barometric sensors with thoughts of adding one to my ad-hoc monitoring rig. I had a hunt around on ebay, as one does, and come up with a little breakout module that has a Bosch BMP280 pressure and temperature sensor on. For the record, it was this one here: http://www.ebay.co.uk/itm/272379854598 and cost a grand total of £2.70 GBP delivered.

I lashed the aforesaid breakout board together with an Arduino Due and and LCD display and for the last two weeks it's been sitting under my computer screen staring me in the face. Once or twice each day I've compared its reading with the Met Office's idea of the current barometric pressure and so far it has agreed digit perfect with the Met Office's idea of what the local pressure has been. The claimed absolute accuracy of the chip is +/- 1 hPa (1 millibar) and that's been bourne out by my monitoring. The claimed relative accuracy is +/-0.12 hPa and I've no way of accurately assessing that but the good absolute accuracy suggests that it'll meet the claimed figure for this too.

The temperature sensor is really there to allow the chip to temperature compensate the barometer part and I didn't really regard this as a possible substitute for the usual temperature sensing one might deploy for boards, references, whatever. For completeness, the claimed absolute temperature accuracy is +/-1 C, and resolution 0.01 C. In practice the temperature readings are a bit high, which I put down to self heating, with a consistent offset from my calibrated K type thermocouple of 1.5 C.

So, if you want a way to log barometric pressure while you're running drift logging on your boards, references, etcetera, it looks like the BMP280 might be a reasonably accurate and very cheap way to do it, as long as you've got a spare I²C or SPI interface handy to hook it up to.

[Andreas]

Hello,

if you go for BME280 (instead of BMP) you have all 3 environment variables (pressure, %rH + temperature).
And obviously the BME280 is also supported by the RaspBerry Pi logging project.

https://www.eevblog.com/forum/metrology/raspberry-pi23-logging-platform-for-voltnuts/msg1120003/#msg1120003

with best regards

Andreas

[azer]

It's performance has been tested rather extensively here
http://www.kandrsmith.org/RJS/Misc/Hygrometers/calib_many.html
and it confirms that some samples have a slight offset in the temperature reading and that the temperature has some measureable humidity dependence.

[plesa]

It's performance has been tested rather extensively here
http://www.kandrsmith.org/RJS/Misc/Hygrometers/calib_many.html
and it confirms that some samples have a slight offset in the temperature reading and that the temperature has some measureable humidity dependence.

Thanks I was searching while ago for some comparison. It just confirmed right choice
I also observed small difference between sensors, but for this price it seems to be unbeatable.

[branadic]

The system I currently use for environmental measurement is SHT25 for humidity and temperature, MS5611 for pressure and temperature as well as TMP112, to measure temperature at certain points. TMP112 is somewhat less in resolution compared to the others, but still good enough.