Using BMP280 as an altimeter for CanSat

Question

Situation:

Me and my student team want to use a BMP280 air pressure sensor connected to an Arduino Nano to measure the altitude of a CanSat.

A CanSat is a coca cola can-shaped and can-sized aerial vehicle, which can be released from a sounding-rocket or a plane-/quadcopter-type drone, packed with scientific instruments to do predefined measurements/operations.

In our case, this CanSat is dropped from 300m hight and has to perform an operation as close to 70m altitude as possible.

Questions:

1. How precisely (practically) can the BMP280 measure the altitude (given physical noise and internal tolerances of the chipset)

2. Where is the ideal spot to drill the measurement inlet(s) into our airframe, when the airframe is shaped like a can and is falling vertically without any drift to the 'sides', and where do we have to position the sensor in relation to the position(s) of this/these inlets?

We would really appreciate the oppinions and practical knowledge of experienced people and enthusiasts :)

Answer 1

Sensitivity

Barometers are quite sensitive, however "air pressure" is very much not consistent. Generally, you can expect about one meter accuracy, maybe a bit less, from a barometer, once it has been given time to settle. This is important to consider if your CanSat is falling quickly, especially if you have small pressure regulation holes

The primary workaround on traditional quadcopters/fixed wing aircraft is to rely heavily on multiple integrations of the accelerometer, and use the barometer as a heavily averaged baseline. This requires more sensors, but you effectively use the barometer to rule out drift in the accelerometer's readings. This is exceptionally complex if you can't predict the orientation of the can (a multirotor will be able to do this math with only one axis), but a vector sum of all axes should get you a close enough approximation

Airflow

As for airflow - you don't want direct airflow to the barometer, don't put it out in the wind. Common practice is to use a permeable foam covering the barometer - similar to how a microphone is wrapped in a very light foam to block pressure waves that will cause interference.

For the can - a similar setup might be ideal. Don't drill a single hole in a single side. You can end up with either high or low pressure depending on the orientation. I would recommend a set of holes on each axis, blocked with foam. This effectively gives you a vector sum, including slight vacuum and slight overpressure.

GPS

GPS provides altitude information. I'm not sure on the accuracy of it, but you can measure pretty consistently at 10hz, which is probably fast enough if you can trust the readings and don't have to average them

Of course, orientation and being in a metal can are issues with GPS