I am building a little and cheap anemometer mainly from scrap parts (pic related). The center is just a random plastic cap, the cups are made of ping-pong balls and the sticks are from a plastic model kit. I am planning to use an old 2½" hard disk drive spindle motor as a pulse output sensor.
Are there any formulæ for estimating the rotational speed vs. wind speed for a 3-cup anemometer of given dimensions? Only information that I've found deal mostly with topics like aerodynamics but the mathematics in those scientific papers is quite involving.
>>945193
IF you can get your hands on a decent properly working one you can run a side by side comparison and than calibrate it. That's how most stuff is calibrated
>>945194
And how the stuff i used for callibration got calibrated?
I mean, what is the absolute method for calibration?
>>945193
you should be able to convert speed of the cups to rotation speed with some maths, very simple.
the point of an analmometer is that its light and very very low torque so it should spin at the speed of the wind.
if you want to calibrate it i would suggest a wind tunnel. they have fancy sensors which find pressure differences like pitot tubes
Have a calm day and take it for a drive. Put it out the window at 5km, 10km, 20km etc.
Use GPS speed as car speedos are often a little off.
Have it far away from the car to avoid the cushion effect
>>945226
>Have a calm day
drive around in a big circle and average out the values to null for air speed
http://mcs.uwsuper.edu/sb/Electronics/Wind/
>>945219
>analmometer
precision instrumentation
>>945195
>measure the same thing n times
>uncertainty shrinks by square root of n
>???
>profit
>>945195
>>945416
What he said.
It's called the least square method.
But it's a bit too /sci/ for me to bother with explaining it. You can google it. But you need to have "standard values" to calibrate to. So like a fan that has an exact speed. And than you have to factor in errors and stuff. If you want to learn about it there as a few books on calibrating instrumentation. But if you just want to calibrate it for personal use just find one that works and run a side by side comparison. assume a 3-5% error on the device and do a graph. If you want it to be a bit more precise do each test 3-5-X times and divide the average. than subtract the average from the value most off. divide by the average value and multiply by 100 you get you % error. take the highest value of error as your instrument average error and boom its calibrated close enough for the girls I go out with.
