DIY Power Meter

[brobee]

So, I would like to get a power meter but I am unwilling to shell out the money (the $200 iSport one is not what Im looking for). Being handy with electronics and computers I started looking into doing a DIY. But I have not done anything mechanical in a long time so I am treading lightly, my confidence is not so high in this area.

So just conceptually I was looking at this as W=VA. Which mechanically, P=TVa (T=torque, Va=angular velocity). Va seems to be simple enough to get. Cadence/wheel speed measurements are common enough. So torque was the problem. And I dont want to build a custom BB or hub to get the job done. So I began looking into other ways to get torque rather than measuring it directly. Wikipedia was naturally the first stop.

And here, http://en.wikipedia.org/wiki/Angular_acceleration, I came across this: T=IAa (I=moment of inertia, Aa=angular velocity). This would be the non-constant case, but I feel that I am missing something here. If you could measure cadence much finer than 1 pedal RPM, then it seems that I could take the derivative of of it (numerically). Is this right? And, at this point... how much does I (inertia) matter? It seems fairly negligible next the force one puts onto the pedals...

[hugofabreu]

I think that by pedal rotation alone you'll never guess if the rider is going uphill and making a lot of effort or downhill and stumping on the pedals.

How about dropping the sensor from a digital kitchen scale between the foot and the pedal?