Motor testing without a dyno.. or even a tach..

[Fumesucker]

So.. I got my brushless motor the other day and wanted to confirm the rather vague motor constants provided by the seller.

The way I go about motor testing is to use two different propellers on the motor, measure the voltage, current and rpm with each propeller at full throttle as well as a no load current reading and then plug the numbers into a motor calculation program (I'm using http://www.peakeff.com/ and like it).

I've had an optical tachometer for a long time and was going to use that to measure the rpms, but lo and behold when I pulled out my tach the batteries were dead because apparently one of my grandkids had been playing with it and left it turned on. Of course the stupid tach uses four watch batteries that if bought locally cost more than I paid for the tach originally.

Since I'm basically an impatient cheapskate I tried to figure out a way to measure rpms without having to spend a chunk on four tiny batteries or waiting to order them from ebay.. After about five minutes of severe head scratching I realized I could possibly do the rpm readings acoustically by recording the sound of the propeller.

My digital camera will take quicktime movies with sound, so I ran my tests with the camera close to the end of the prop recording the sound of the blade tips going by. I then extracted the audio from the movie and put it in an audio editing program and was gratified to see that the blade tip passing by the microphone was conspicuously obvious.

Any method of recording audio will do, just so long as you can get the mic near the blade tip.

The freeware audio extraction program I used is here: http://www.aoamedia.com/audioextractor.htm .

The freeware audio editor I used is here: http://audacity.sourceforge.net/download/

And here is what the recording looked like in the audio editor..

So I then used the spectrum analyzer built into the audio editor to find the frequency of the tip passing..

Note in the spectrum plot window on the bottom there is a readout for cursor and peak, the peak is at 79 Hz, which means the blade tip was passing at 79 times per second.. (79/2)*60=2370 rpm..

There you have it, rpm measured acoustically and all with freeware..

 

[TylerDurden]

That's excellent...

I rekon the same technique could be used with wheels and cogs, by putting a flexible tang (like a playing card) against the spokes or teeth.

 

[Fumesucker]

I hadn't thought of that, TD, you just need to make sure that the sound you are using to measure rpm is the loudest one by a fair bit. It also helps if the sound is the lowest frequency among the sounds the mic picks up.

Milk jug material works pretty well for shade tree guitar picks, I don't see why it wouldn't work for the purpose you described.

I'd also like to point out that you can measure the rpm just by looking at the scale above the waveform, it is calibrated in seconds and fractions. You can measure the period of the waveform directly this way if you have a waveform that is less obvious in the spectrum analyzer plot.

 

[Ben]

Nooooo, you stole my idea!

Well, actually my idea was to use a playing card against the spokes and record the audio, then convert it into displacement vs time, velocity vs time and acceleration vs time graphs.

 

[Fumesucker]

Now that *is* a good idea.. I definitely like that.

My cell takes voice notes, that would be more than adequate to record the audio from the spokes..

If you were to use a digital video cam you could vid the wattmeter at the same time and get some really comprehensive numbers.. Speed, acceleration, voltage and amp draw and you could get some pretty decent numbers on the total system efficiency..

 

[Link]

ROFL

I used a similar method for figuring out how fast some of my airsoft guns could shoot. Shot through two pieces of paper 1ft apart and picked it up with a pair of headphones because the only mic I had had broken.

I need to figure out how fast my new gas guns are, now that I think about it...

 

[mcstar]

Wow, this sounds like something I'd have done. I'm still ROFL! To the non-tech I'm sure it seems a bit Rube-Goldberg, but hey, use what's available (for free) right!
http://mousetrapcontraptions.com/

 

[John in CR]

Great idea Fumesucker. Now I've got to load my test tone generator in the laptop. It should be quite easy to dial the rpm right in by ear by varying the tone rate while running the motor making a sound with each revolution.

John

 

[Fumesucker]

John,

Keep in mind that the waveform is very harmonic rich, in fact when I looked at it with SpectrumLab the first harmonic is actually several dB louder than the fundamental frequency..

Which means you should start with a low frequency and sweep up until you get a beat with the prop sound.

 

[Mathurin]

Nooooo, you stole my idea!

Well, actually my idea was to use a playing card against the spokes and record the audio, then convert it into displacement vs time, velocity vs time and acceleration vs time graphs.

I recall playing with software like that for cars a few years back, you put the number of cyls, weight and gearing, and it gave power curves pretty close to published specs. Something like this could be used as cheap dyno for testing max power in australian class ebike racing.