Sound Energy to Electrical Energy

Connect an LED to a mic and yell real loud. Not a new concept.

If we give you the winning idea, will we get the prize?

Here's one. Hook the magnet to an electric vibrator. Of course, the vibrator will consume far more power than your contraption will produce.

What frequency?

Even a speaker coupled to the vibrating plate will produce a current, however, the amount of energy produced is not going to be much.

I would start by trying a large woofer and mechanically couple the voice coil to the plate. The magnet structure of the speaker must be mechanically mounted to something that is solid and does not vibrate.

The output of the speaker terminals will produce an AC current proportional to the magnitude of the vibration. You may need a substantial vibration to drive a LED directly.

Another solution is a piezoelectric sensor, but the output is going to be very low and would require amplification.

Instead of LED lighting, Charging cellphone battery is a better Idea when you are dancing on DJ.But make sure DJ owner don't know it.

I think you would need to tune the system so it resonates at the frequency of interest. It's then about getting the the magenetic field to cut across the maximum number of turns of wire. As soon as you try to take power out it will damp down the resonance, so the whole mechanical and electromagnetic arrangement will need tuning to get the best output. Thus it is best if you are trying to harvest power from a known input, say a machine running at a steady rate.
In fact I'd try to harvest the mechanical movement which causes the sound rather than the sound itself, cut out the middle man.
Del

Can u sugest some ideas to present for an inovative engineering contest?

It sounds like you have connected an LED, some diodes and a capacitor to the terminals of a speaker and are running it backwards to create a simple DC power supply powered off an alternating current source. This doesn't sound very innovative for an innovative engineering contest.

Have you considered the developing engineering field of vibration energy harvesting using piezoelectricity?

Link

Can you please provide more information. Additionally what is the project for, is it a school project, university engineering project, etc.

Loudspeakers have the strong magnets etc you need, but their efficiency to sound is poor. I believe Horn loudspeakers are by far the most efficient. Also, acoustic phonogram and disc systems had to have big horns to get enough power to just scratch a groove in a disc. In short, you must have good acoustic matching to get much power. Think 10 metre parabolic concrete "mirror" with air-raid siren in front of it.

Further to post #9, I looked up the fundamental numbers.

A really loud sound - underground express train passing station - is 100 dB [ 0 dB is 10^-12 watts per square metre] and that is acoustic energy of 1 microWatt/sq. cm. Hence 1 square metre at 10% efficiency would give 1 mW. One milliamp at 1.5 volts DC should give visible light from an LED (there is a minimum voltage and a quick trial with a small red LED was clearly lit at 0.1 mA). For good effect, I would think in terms of energy storage in a capacitor and a flash circuit. As has been commented, having sound of a definite frequency would help, since tuning the system would help efficiency.

Oh dear, the usual.
Plenty of input and questions from us.
Zero feedback or answers from the OP.
Maybe Fredski was right?
Del