Help with sound sensor

[tee_2]

Hi I need some help how to do this if possible. I want to make some kind of sound sensor that will pickup the sound of a small steel ball hitting a steel plate. Some times the ball will hit soft (low sound) and some times it will hit hard (high sound). I need to trigger a LED that will come on for 15-60 seconds. I can handle the LED part of the circuit using a 555 timer.

Now for the hard part. I need sensor A to not be triggered by a sound that is meant for sensor B, that might be 3-6 feet away. If I can do this there may be up to 20 sensors, but no closer than 3-6 feet apart. Would I be able to use something like this https://www.jameco.com/webapp/wcs/s...toreId=10001&catalogId=10001&productId=121252 and put it in a tube to make it more directional? The sensor can be two, to 12 inches away on the back side of the steel, or three to five feet in front of the steel. In front is the better of the two for mounting the LED.

Any Ideas? Keep in mind I have only been playing with electronics for a few months, so I don’t have much of a clue. But I can make a LED flash using a 555 timer.

Todd

 

[tee_2]

No replys? :cry: Does that mean what I was thinking can't be done? Anyone know a site that would show how to make a simple switch that uses sound?

Todd

 

[williB]

https://www.electro-tech-online.com/custompdfs/2005/02/LM311.pdf
you could connect a microphone to a comparator and send the comparator output to your 555 timer ckt..

 

[stevez]

Will the impact, both soft and loud, produce a similar sound? Will the energy/output of your microphone or other sensor be much higher than other background noise when an impact (soft or loud) is made? If the loudness or amplitude of the impacts you want are signficantly different than the other noises, including those that you don't want to trigger the LED, you can use this to advantage - with simple comparators or other triggers. If I were doing a project like this I'd do some background experimentation first to see what you have to work with. Other things you might discover is that the hits you want to record might be followed by a distinct ringing (the metal plate is likely to 'ring' after a hit, regardless of how hard it's hit). You could work to be sure that different plates ring at different frequencies. Some simple audio filters might help you sort out the impacts you want from those you don't want. I am sure there is software out there that could be employed but you probably want to keep it simple.

 

[audioguru]

Hi Tee,
I think that your sensors are so close together that a loud sound from a nearby machine will trigger the local sensor when the sensor is sensitive enough to trigger with a low local sound.

Don't make the microphone directional with a solid tube, it will resonate. Use a tube made of thick foam rubber with dense stuff (a plastic sewer-pipe would work well) around its outside. Put the mic in the tube near the distant end with lots of foam behind it. Then the mic will be directional at high frequencies if the tube is long enough.
The parts around where the steel balls drop will reflect sound, so should be well-padded.

Of course, the activity you want to detect is audio, so will be interfered with by any other audio like car horns or planes flying over.

The electronic circuits that you need are a preamp for the mic, audio filter and a rectifier/filter that will trigger your 555 timer.

The mic that you posted is awful darn cheap so might be a factory-reject and if they work, will probably all be different. Get good prime mics from Digikey or Newarkinone (Farnell). One of these suppliers has 30 mics to choose from.
Good luck.

 

[Oznog]

You could use a differential mike. You'd put two mikes near each other, maybe 1/2"-2" apart, and put them into a differential op amp or just put them in series with one of them wired in the opposite direction.

The key is when they're far away from the source, position has little effect on the signal, so the voltages cancel out and there is no input. Nearer the source, the amplitudes between the mikes will be different thus a signal results.

Bottom line is it will be much more insensitive to sounds originating further away. Of course a mike being closer to the source gets a higher amplitude too, but this effect of this arrangement is stronger still.

A mike could physically be places on the metal too, if your project allows. The strength of the signal will be way stronger than any other sound.

 

[pebe]

Tee_2. You have not said what the application is, which may have helped.
I would have thought a steel ball hitting a steel plate would make a sharp 'ping' sound, rather than a 'soft' or 'hard' sound. Do you mean 'quiet' or 'loud'?

I think the sound produced will have a sharp leading edge which could be differentiated to produce a pulse. That would get rid of extraneous noise. An electret mic fitted close to, or even on, each plate would give a good response with minimum pickup from a neighbouring plate.

The individual mic outputs could be fed to a device that detects which mic is being actioned and the signal proccesed from there

 

[jbeng]

A piezoelectric transducer could be used in a manner similar to what pebe described. Mount it to the back of your steel plate. When the ball strikes the plate, you should get a pulse from the piezo which you can then amplify and send to your logic circuitry.
JB

 

[FRIED]

Circuit from a project to pick up a sound and output a pulse. C1,R1 set pulse length.**broken link removed**