Estimate Mass using Accelerometer

I would like to be able to strap some sort of vibrator and accelerometer onto the side of a hanging mass. By imparting known forces with the vibrator and reading the side-to-side acceleration I would like to calculate the mass of the hanging object.

If I were to use a motor with an offset weight on the shaft it would be imparting a sinusoidal force, so the acceleration would be sinusoidal as well and I think it would be phase-shifted. I think the waveform amplitude and phase math would be similar to an AC impedance/voltage/current circuit. But I don't know much about that either, especially how such would be read and calculated by a microcontroller (Arduino) in "real time". Or is there some AC circuitry or DSP that could do the work of the waveform analysis to present the arduino with a signal which was directly proportional to the unknown mass, e.g.

One major reason I am seeking to do this in this less-obvious way, is to learn more about practicalities of AC and signal processing in an Arduino context. Maybe a similar question would be: "How to use an Arduino to measure capacitance?".

I thank you for any thoughts you may contribute!

Cheers, John

The measurement could be done with a Newton Pendulum: - http://www.youtube.com/watch?v=mFNe_pFZrsA -

To measure the unknown mass you first let 1 ball bounce, then 2 , 3 , 4 , 5. For every number of balls you measure the speed the unknown mass gets. This gives 5 measurements which can be averaged or so.

For the know balls you can calculate the speed of impact as it depends on the heigth you lift the balls:

Potential Energy = Kinetic Energy => mgh = 0.5mv^2 => v = sqrt(2gh)

And then apply the law of conservation of impuls: m(known) * v(known) == m(Unknown) * v(to be measured) ==> m(U) = m(K) * v(K) / v(measured);

and because v = a x time you can measure the speed with the accelerometer.

In short, it could be done , don’t know how precise it all will be.

An Arduino controlled balanced scale could be easier ( a robotic arm putting on weigths in one scale and checking the tip :wink:

sofar my 2 cents,
Rob

I like your 'eccentric' motor idea. You could take one out of an electric toothbrush. I'm not sure why you need the acceleration, surely all you need to know is how far it moves on each cycle. Lower mass - moves further.

You could use a Sharp IR distance meter module. Although this may not be accurate enough.

Si: I like your 'eccentric' motor idea. You could take one out of an electric toothbrush. I'm not sure why you need the acceleration, surely all you need to know is how far it moves on each cycle. Lower mass - moves further.

You could use a Sharp IR distance meter module. Although this may not be accurate enough.

I was thinking that the accelerometer would allow a more self-contained package.

I don't know if the resolution (and/or "frequency response") of an accelerometer would be adequate either, at the scale of a toothbrush motor (or small toy moter either).

Cheers, John

DO you have a rough estimation of the mass? order of grams or kilograms?

robtillaart: DO you have a rough estimation of the mass? order of grams or kilograms?

Oh, good point! 5-10 kg

Thanks, John