With how common the common case fan is with a tach output, I thought it might be possible to make a anemometer from one. So I hit google and came up with nothing. My knowlege of fluid dynamics is basic, and I'm not sure what to search on to find the math I need. The plan I had was to take the fan off the hub and disconnect the coils allowing the fan to spin freely. Then conect the fan to either a mainboard header, or use the maxim tach IC with I2C output to get readings off the fan. I could then write a program to convert to air speed. I thought it would be a fun thing to try to build and use it to test fans, HSF combos (sort of an airflow efficency test for heat sinks), ect. The electronics aren't a problem, its figuring out the relationship between the RPMs of the fan and the air speed. What should I be looking for?
well, first off, you will have to take the magnets out of the fan - they offer too much resistance. then your best bet is to find some fans that you know the CFM's on and hook them up. speed is a lot harder - you have pressure to worry about.
Problem, the RPM signal in all brushless fans i have seen rely on the spinning magnet to provide a magnetic field to a small magnetic sensor.
i was talking about the electromagnets! you are right, you need to leave the permenent magenet in - it is very hard to take out.
If you have a UPS or a DC/AC converter for a car you could try this http://www.alphalink.com.au/~derekw/ane/calib.htm - of course you need someone with a car too.
I got that part already. Most of the fans I've taken apart are basically an RC circuit with a hall effect sensor to sense the RPMs, and help regulate the speed. I could try and get the fan far enough apart to fiddle with the RC circuit, but this is usally bad, I've seen too many fans that have the circuit board locked to the hub somhow (epoxy?). But I can get at the coils in nearly every case. Then all I should have to do is supply power and read the tach signal. I'll have to try the car calibration method, but I think I'll need an I2C to parallel or serial reader. Then I hope I'm set. I'm pretty curious as to how varous heat sink designs affect the rated flow rates of the fans mounted to them. Thanks for the link nick01
It's not quite a RC circut. The coils are actually driven by the hall sensor through a transistor. But, yes. If you tear the coils out on a Ball Bearing fan, it will spin quite freely.
I have to look at another fan. There needs to be some sort of DC to AC conversion or some similar oricess for the coils to energize for the right amount of time for a certian rotational speed. But I traced out the schematic before I knew that the black thing was a hall effect sensor. I'll look again, because another project that might be fun is to 'overclock' and 'underclock' a fan by messing with the timing circuit. (if it is the main governing circuit for speed control)
There really is no way to over or underclock a fan circut, rather than actualy changing the gauge of the wire used to wind the electromagnet (or the voltage; You should see a 12V fan at 24V It doesn't live long, but it actually hovers when placed aainst a flat surface). It's resistance determines the ammount of current (read: enenrgy) flowing through the circut. You might be able to modify the speed of the fan by rotating the pcb WITHOUT moving the coils, but anything that made it go faster by that method would also likely make it not start from a dead stop.
On a standard fan, it's simple commutation as on any brushed DC motor; the Hall sensor lets current through to the "right" coils when the rotor is in the "right" position and the permanent magnet's poles activate it. Once the rotor's moved on those coils turn off. http://www.cpemma.co.uk/brushless.html
This is kinda relevant to fan hardware, you'll need to sign up for OCAU but it's generally worth it. http://forums.overclockers.com.au/showthread.php?t=380518
I have seen some fans that do appear to use RC circuits to energize the coils, but after taking a few of some newly worn out fans apart, I do see some that might be dubbed 'inertaly regulated' as the rotating mass andair resistance of the bladed seems to be the main factor govenering speed. So there doesn't seem tobe any one system. Thansk for that info cpemma and Wolfe. I'll have to sign up and check out the site you linked to g0th. I have so many fans lying around, I want to experiment with them like the anemometer. As for speed control, PWM and voltage control seemed inelegent, at least for the ones I have taken appart and found some control circuitry. But these really simple ones, those seem to be fine. I still think I'm goind to experiment, but I think I'll try a new approach and make my own inverter to feed the coils directly with the hall effect sensor helping to regulate factors such as speed ramp up, and seeing if the fan is at it's max speed.
