signal waveform feedback

[Lorentz]

Our experiment is as follows:

We send a sine wave to a nonlinear speaker, which in turn gives a sine wave with extra harmonics.

The displacement of the speaker will be measured with accelerometers.

The acquired signal from the accelerometer will then be used as a feedback input to compensate for the 2nd and higher harmonics.

If this compensation is done in the right order, the resulting signal from the speaker must be a sine.

We have tried a lot but we are not able to compensate in such a way that the speaker will give a sine.

Is there a standard routine to handle such problems?

Thanks

A.

[johnsold]

Does your feedback work fast enough to make the corrections in real time?  Or are you calculating corrections one time and then applying them in hardware?

Please describe your configuration in detail.  How are you acquiring the acceleration signal?  What is the range of fundamental frequencies? What is the sampling rate?  What is the resolution? How do you generate the "compensation" signal?  How is it combined with the original sine wave?

Have you tried to model your entire system as a closed loop feedback system?  Does the model, with realistic parameters, work?

Lynn