I'd like to build an airflow sensor (for low pressure, max. 200 Pa and a flow of max. 1 liter/s), but I don't know what would be the best way to build it, considering there are many different possibilities.
First solution that came to my mind was to build something like a rotameter, which could be similar to a one-way valve (a moving inner part in a conical tube) where the inner part of the valve would be able to move in a linear direction and would be forced against the aiflow by a small spring (meaning that when there would be no airflow, the "valve" would be closed and when there would be a big flow, it would be open to the maximum). Then the position of the inner part could be translated to a potentiometer and determined by the value of voltage drop present on the potentiometer (but I'm afraid the friction would be too high) or to a small plastic sheet with a triangular hole in it - the sheet would be located between a LED or phototransistor (or a small PV cell for example from a calculator) and the amount of light passing to the phototransistor would be proportional to the sheet position.
I've also found a method that is based on using a heated element (piece of wire, a diode or lightbulb with the glass bulb removed) - called "hot wire" - and measuring its change of resistivity depending on the air flow.
The first option seems more straightforward to me but would be probably a little bit complicated to build, and I'm not really sure about the precision (I'd like to take samples of the air-flow values and use them to calculate an overall volume that passed through the airflow meter) and maybe also the inertia of the inner (movable) part could be problematic (possibly the measurement could be influenced by movement of the whole airflow meter and response to changes in airflow could be slow to reflect fast changes).
The other option seems more sophisticated, but I assume that the resistance change would be quite small so amplification of the sensed voltage (to determine the resistivity) would be needed, which could be more prone to errors caused by noise (or not? - I don't know, I have very little experience with electronics in practice) and also it could be influenced by the temperature of the air passing through (probably this could be overcome by doubling the measurement on two different places and comparing it or something like that - but I don't have a precise idea how to do it).
I'd be glad for any insights in this area or any hints!
Using a spring has a problem- springs change force with temperature.
A hot wire wind speed indicator isn't too hard to implement. What you do is build it as a wheatstone bridge, with another hot wire that is shielded from the airflow as the second arm of the bridge. It is called a hot wire anemometer. The usual version uses a very thin platinum wire and is rather fragile. but responds very quickly to changes in air or fluid velocity. It can be incredibly precise, but it is affected by temperature changes unless you compensate for them.
It could be as simple as two silicon diodes that have sufficient current flowing to heat them. One is in the airstream, the other is inside an insulated container that is in the immediate vicinity of the first diode. A differential Op Amp configuration handles offset and amplification.
In fact, TI has several ADC chips that contain PGA (programmable gain amplifiers) and 12 or 16 bit ADC, and can be configured as differential inputs. Adafruit sells the 16 bit ADS1115 on a breakout board for $15:
I'd like to build an airflow sensor (for low pressure, max. 200 Pa and a flow of max. 1 liter/s), but I don't know what would be the best way to build it, considering there are many different possibilities.
liters is a volumetric measurement. velocity is what is measured.
do you have an accuracy you need too try to obtain ?
polymorph: I think temperature changes wouldn't be an issue with the spring, because the temperature wouldn't nearly change at all for it to change mechanical properties. The "hotwire anemometer" looks interesting but I'm afraid I don't have enough knowledge yet to put together a working one...
dave-in-nj: Of course I know that velocity is measured, but AFAIK when I know the diameter of the tube the air is passing through, I can calculate the amount of air that passed through it based on the velocity at different points in time. I don't have any exact needs on the accuracy since I don't know how accurate the measurement could be considering the measurement tool shouldn't cost a fortune (something like $25 max - of course with the price of arduino excluded)... the measured volume could be from something like 5 litres to 50 litres (the goal is to measure how much air different kinds of balloons can take), so maybe 0.5 l accuracy would be good enough? Though I don't know if it is achievable with some not-at-all sophisticated equipment.
