The flow sensor is made of transparent plastic case of one cheap analog alarm clock.
When I removed the clock mechanism, I got a box, circa 2.5" x 2.5" x 1.5", all I needed
was a back plate. The walls of the case are 4 mm thick so there are no fear of cracks and leaks
(hitting it with hammer excluded...
). Backplate is simple, piece of plexy cutted to fit
tight on the back of the clock case. I disassembled one broken 5cm fan, took EVERYTHING out,
leaving only plastic holding a sleeve bearing, also took out magnet strip from fan propeller.
You must rip out everything to make it lighter and easyer to spin, and to prevent bubbles of air get stuck inside.
The ripped fan is then epoxied to the backplate, and backplate epoxied to the back of the clock case.
The barbs are also epoxied (NO goop of any kind to buy here, grrrrr), the only thig one must
be careful is when drilling the holes for barbs not to scratch the front plate with drill bit
(been there, done that...twice...
). Test it, and that's it!
The idea is to make a optical counter (like in mouse) to watch the flow and trigger the alarm
if necessary. I just couldnt find the way to use magnet and Hall efect (I think...) device
already in the fan (plastic walls being thick, long distance between magnet and sensor),
so I went to make it optical.
Still looking for a good (and simple) schematics.
Im aware that this flow sensor is unnecessary and adds 2 more 90 deg. turns to the flow,
but I couldnt help it...
Geee, this sounds a bit like a doctorate, but sorry guys, I'm not native english speaker.
No time toulouse.....Monty Python...