Fan fun
December 24, 2012 8:41 AM

How do I, using common household tools and objects, determine the (variable) RPM of a fan?

To set the scene, we recently put in a woodstove that we're using to (mostly) heat the house. This stove combines two wonderful things: 1) obsessive experimentation with a multitude of variables that affect its efficiency, and 2) playing with fire. So this is but one of the small facets of its operation upon which I'm nerding out.

The fan atop my woodstove is powered by a Stirling engine. I move it around to optimize its speed and direct more warm air in specific directions. The fan speed is determined by the temperature differential between heat conducted from the stovetop and cooler air passing over the fan's cooling vanes. I can determine the hottest point on the top of the stove with great precision, but because the fan speed is also dependent on the flow of cooler air reaching the cooling vanes, it's difficult to determine which placement is yielding the highest RPM. The fan is spinning too fast to judge its speed.

I would like to determine the approximate RPM of this spinning fan under different conditions. I thought about taking photos of the turning blades with different shutter speeds on a dSLR, and comparing the blur, but the pre-programmed shutter speeds seem too blunt an instrument for the measurement I'd like to take. I thought about using something that flickers at a defined rate, but I don't have a CRT and can't think of anything similar.

Home engineering types, can you cobble together a means to take an approximate measurement and further enable my stove tinkering?
posted by itstheclamsname to Home & Garden (17 answers total) 4 users marked this as a favorite
Attach a pencil or something to the end of one of the fan blades. This may throw the fan out of balance so you may need to attach other items of similar weights to the other blades to rebalance it. Then put a playing card or zip tie or something on the path of the pencil, so that for each revolution, the pencil flicks past the card, like a baseball card on the spokes of a bicycle. Record the sound and count the ticks per second. Slow it down on a computer if need be. I bet this works up to at least several hundred rpm before the ticks turn into a hum.
posted by tylerkaraszewski at 8:46 AM on December 24, 2012


I was going to suggest the same thing. Use Audacity to expand the waveform and you can see exactly how many ms are between each peak. Divide that into 60,000 and you have the RPM at a given moment.
posted by griphus at 9:09 AM on December 24, 2012


i love this question...

I haven't yet solved the precise question, but i have solved the root measurement issue.

1. get some length of fishing line or very fine thread. (mark a "start pont after some amount of thread and a "stop point" later on on the thread with color or a knot etc)
2. attach the thread to the shaft of the fan.
3. start the fan
4. time carefully how long it takes to wind up the thread from the start to the stop mark.

this will NOT get you the "RPM" per se but will give you a measurable repeatable way to determine how fast the fan is spinning and if there is a change in rate. change the fans location and repeat the test. did the string get taken up faster or slower?

I would probably test it multiple times at a given location over some repeatable amount of time and then calculate the mean and standard deviation. i would probably due this because i'm a nerd. here's a calculator to help.
posted by chasles at 9:12 AM on December 24, 2012


Buy a bike computer and attach the magnet on the blade (counter balance the other blade).

Sorry in advance if this is not cobbled together enough for you…
posted by Rabarberofficer at 9:14 AM on December 24, 2012


I thought of a (relatively) easy way to do it with an arduino, too, but I don't know if an arduino counts as a common household object. You could get a little LCD to just display the RPM count, though, which would be cool, and you wouldn't need to add hardly any weight to the fan itself, avoiding the balance issue.
posted by tylerkaraszewski at 9:23 AM on December 24, 2012


Minor improvement to the audio method -- make the clicker something lightweight (like a toothpick) and mount it near the hub. It'll affect the balance less and have less mechanical advantage vs. the drive shaft, so the drag from the clicks will be reduced.

Taking pictures at a known shutter speed and measuring the blur is actually a time-honored method for figuring out fan blade speed. It helps to make one blade high-contrast against the background so you can see exactly where the blur starts and stops (so mark up a blade w/ black sharpie or white paint/white-out). Having one blade a different color than the rest also means you can handle a whole rotation's worth of blur; otherwise all the blades will smear together. If you can put the camera on a tripod (or just nail it down somehow), you can take one picture with the fan still to use as a measurement reference so you don't have to work out any geometry or pixel-width math; you'll be able to say that the blur is X blades wide. As long as the shutter timing is relatively accurate, it shouldn't matter that you don't have a lot of choices. If you can find 2 or 3 settings that are good (the trick is harder if the blade moves too far, or not far enough, in one exposure) that should be plenty to verify and get a handle on your shutter accuracy.
posted by range at 9:36 AM on December 24, 2012


Oh there is a ton of fun ways to do this with an Arduino, of varying levels of practicality. If you can shine a light behind it, you can use a light-dependent resistor to count each time a blade cuts the light.
posted by griphus at 9:52 AM on December 24, 2012


Doesn't directly answer the rpm measuring question, but:

You could make a windsock and attach it to the front of the fan. After all, fan rpm is just a way to guess what you really want to know, air flow rate.

(Which is just a way to guess what you REALLY want to know, rate of heat transfer from stove to bulk air in the room. )
posted by ctmf at 9:55 AM on December 24, 2012


Does the fan make noise, clicking of the piston or anything? you could record the audio of it and count peaks in audacity or something.
oops, sorry griphus
posted by duckstab at 10:19 AM on December 24, 2012


If you can paint a bright or reflective dot on one blade, the camera idea should work well as long as you can get the blade to go through some significant fraction - but <1 - of a rotation. You could even try to put degree marks around the edge.
posted by ftm at 10:22 AM on December 24, 2012


Record yourself humming at the same note as the fan. Then look up the frequency of that note. If the fan's note is too low or two high, aim for the same note in a higher or lower octave.
posted by zippy at 11:09 AM on December 24, 2012


Can you find a strobe light with an adjustable speed? If you match the strobe speed to the fan speed I believe the blade will appear to stop moving...
posted by NoDef at 11:11 AM on December 24, 2012


If you are going to buy anything a hand held non-contact digital photo-tachometer can be had for as little as $15.
posted by Mitheral at 12:25 PM on December 24, 2012


If you can shine a light behind it, you can use a light-dependent resistor to count each time a blade cuts the light.

A ball mouse is essentially this plus a USB encoder. One might adapt an old mouse to the task and write a bit of software to grab the pointer and measure its motion over time. It would require calibration, however.
posted by rlk at 1:24 PM on December 24, 2012


You might be able to use the rolling shutter effect on a camera phone to estimate the RPM of the fan. For best results you would need to (a) know how long the exposure took from start to finish, and (b) have fan blades of a simple shape. You could get around the latter requirement by, say, putting a straight line of masking tape on the fan blades.
posted by Johnny Assay at 4:23 PM on December 24, 2012


Hum a single tone from behind the fan and place a microphone in front of the fan. Use audacity software to visualize the recording. You should be able to get at blade RPM by counting the # of volume modulations per unit time.
posted by surfgator at 3:03 PM on December 25, 2012


This is far more complicated than the other suggestions already made, but in the interests of "geeking out"...

Attach something that will make a continuous sound (a whistle?) to the tip of one blade, add equal weights to the tips of the other blades for balance. Make an audio recording from a position edge-on to the blades, figure out the shift in frequency due to the Doppler effect between when the whistle is moving towards you and away from you. This will give you the speed of the fan blade tip; from this you can work out the RPM.

Of course, once you've gone to the trouble of making the recording you could just work out the time between Doppler shifts which will give you the frequency directly.
posted by primer_dimer at 3:24 PM on December 26, 2012


« Older Reading Aloud Allowed?   |   Is SoundCloud a good gift for a new electronic... Newer »
This thread is closed to new comments.