Suggestions for a very precise absolute barometric pressure sensor ?


I am looking for a precise absolute barometric pressure sensor, the price and wiring system is secondary . (with breakout board , read out should be digital, and there should be a library for arduino available)

The accuracy should be better then 1 hpa in the temperature range ~ 0 - 20 °C & 300 - 1100 mbar/hpa.

As there are a lot of (a.e. I2C) sensors on the market with different datasheet specification, which don't always show the min or max error for the absolute pressure accuracy (just typically error ) this is to unsecure for me needing.

At the moment I am using the Bosch BMP180 but the min and max error is around -4 hp to + 2 hp which is to large , the a.e. Bosch BMP 280 got a better accuracy but actually I didn't find a manufacturer who is selling this sensor with a breakout board.

I would like to emphasize I am not interested in the relative pressure accuracy or on the resolution, just on the absolute pressure.

If someone got experience with other sensors which may fit, any help is welcome .

Thomas states The BMP180 offers a pressure measuring range of 300 to 1100 hPa with an accuracy down to 0.02 hPa in advanced resolution mode.

so if you set in in Hires... could solve your problem..

the Bosch BMP180 but the min and max error is around -4 hp to + 2 hp which is to large

Do you have any way of actually detecting that the reading is too high or low? In other words, what are the consequences in your project if the pressure is wrong by, say, -2hp?



I know that sentence and its confusing as it has nothing to do with the absolute pressure accuracy .
Thx anyway :slight_smile:

From Bosch BMP180 Datasheet

Absolute accuracy

VDD = 3.3V

300 . . . 1100 hPa 0 . . . +65 °C Min -4.0 Typ -1.0* Max +2.0 hPa


The absolute accuracy specified by the manufacturer specifies the deviations of the correct measurement for a random selection of devices. If you have the capability of calibrating the device (which you MUST do anyway if accuracy is really important to you), then what matters is the long term stability of the measurement.

In other words, an individual sensor may perform much better than those specifications imply -- you just have to calibrate it.

Out of the box, uncalibrated, it will be between -4 and +2 hPa. Putting that another way, they do not perform an exhaustive calibration of the chip before it leaves their factory -- they leave that to you.

The precision is .01 hPa, RMS noise as low as .02 hPa and it states it won't drift more than 1 hPa over 12 months.

I am using BMP280 for several weeks now and they are quite good!

I use Digikey's proto board 33608CA-ND for BMP280 and it feels like the board is made for it, extremely easy to solder by hand.

All the best, - Aditya