I have a selfbuilt ventilation system in my house, including a selfmade heat exchanger.

Of course I'm interesting to measure efficiency of the system, so I added temperature and humidity sensors to it and even realized a function that it only ventilates when the outside humidity is low enough to avoid condensation at cool walls.

But are the in- and outlet ventilators running at the same flowrate?
How big is the pressure drop of the piping and the corresponding reduced flowrate?

Finally I decided to build flowmeters.
I did not want to introduce additional large pressures losses. These would be generated by putting a "sensing" fan in the pipes.

A good principle is the Vortex flow meter principle. A nice video with a description can be found here:
http://www.youtube.com/watch?v=GmTmDM7jHzA

A bluff body introduces alternating pressure signals after it. The frequency of the alternation is approximately linear to the volume flow rate.

I found a very cheap sensor for these pressure signals: Simple electret microphones.

I had a few drain pipes laying around and decided to built a test-rig by glueing a ventilator to it:



The bluff body is just a small stripe cut from a copper cladded PCB:



I put 2 microphones on the back-side (downstream) of the bluff body:



And wired them to a small breadboard circuit:



Why I used 2 microphones? A microphone will pickup each sound and not only the vortex signal. Especially vibrations of the pipe can be seen very well and can get larger as the actual signal.

When using 2 microphones, the output signals can be substracted. As the dynamic pressure signal has a different sign on each side of the bluff body, this will increase the actual vortex signal amplitude. Signals that have the same phase on the 2 microphones will be canceled out by substraction.

Here a output signal of the circuit when the ventilator is running:



The upper signal is the output of a single microphone. As it can be seen, the vibration of the pipe generates a high frequency component.
The lower signal is the difference signal of the 2 microphones. The pipe vibrations are not present anymore.

This measurement principle works very well, even with low flowrates. The lowest visible flow was 10m^3/h on a DN125 pipe.

Next steps:
Build a circuit to measure the vortex frequency, calibration, etc...