In Vino Veritas - a Wineglass Oscillator

After I had finished a tuning fork oscillator, my brother challenged me to make an oscillator using a wineglass.
( https://www.instructables.com/id/Tuning-Fork-Osci... )

He thought it would be more difficult to use a wineglass than a tuning fork as the frequency determining element. It is.

Everyone knows the sound a (wine) glass makes when you gently tap it, usually it sounds like a fast decaying "ping". Some, more expensive glasses can keep "singing" when you rub a wet finger over the edge. The sound this produces is caused by the glass quickly vibrating in a special way. The round shape of the glass changes into an ellips, back into a circle and then into an ellips but rotated by 90 degrees, and so on. The air vibrates with the glass and a tone is the result.

You can even find serious research on the vibrations of wineglasses, just Google for : "a study of wineglass acoustics" and see the pdf below. (I admit I didn't read it all)

Step 1: Making the Wineglass Vibrate

When I build the tuning fork oscillator, making it vibrate was easy, you just have an electromagnet repeatedly attract it. But with glass magnetism not an option. I could have made a contraption with a mechanical wet finger, constantly rubbing the glass. But mechanical solutions aren't really my strong suit.
Then I thought of attaching a piezo element (as you can find in "musical" picture cards), but I didn't like the idea of anything touching the glass. And it would change the natural frequency of the wineglass too.

It is possible to make a wineglass vibrate with soundwaves. I guess everyone has seen movie clips of wineglasses being shattered with powerful soundwaves. I didn't need sound that powerful, I thought... So I chose an ordinary loudspeaker to produce the soundwaves that makes the glass vibrate.

Step 2: Detecting the Vibrations

An oscillator needs a closed loop, so I had to register the vibrations, amplify them and feed them back (with the correct phase) via the loudspeaker to the wineglass. How to detect those vibrations. Well that proved to be the most difficult part.

On TV I have seen guys working for "three letter organisations" listening to the vibrations of window panes that in turn were vibrating because of voices in the room behind it, with what is called laser-microphones. I thought it wouldn't be all that difficult to make such a device myself as the glass I am listening to is just a few millimeters away as is the laser.

I was wrong. Those laser microphones use the interference of the original laser light and the reflected light to detect the vibrations of the window panes. I can't think of any way I could make a device to do that. Maybe someone else here does, please tell me in the comments below.

Using a microphone to listen to the wineglass doesn't work either, the sound coming from the loudspeaker will be stronger and the system will oscillate, but not with the frequency of the wineglass, you possibly know the squeal when someone turns the amplifier up too much and that sound returns via a microphone.

With the tuning fork oscillator I used an optical interrupter to detect the vibrations of the tines. That worked well, could I repeat that with something made of glass?

Glass bends light, maybe that could be used. So I tried with leds of different colours shining through the wineglass in different ways and detect any changes with a photo transistor. It didn't work. Then I tried a laser light beam reflecting off the glass and trying to detect any vibrations in it. That didn't work either.

What did work was skimming the laser beam across the glass in such a way that the wineglass would block most of the light, the light that reaches the photo transistor is modulated with the vibrations of the wineglass. The trouble with this setup is that it is extremely sensitive to the smallest of movements of the laser, the glass and the detector. But it is the way I made it work.

Step 3: Green Lasers Are Dangerous

First I used a green laser as I know that green laser light is made with an IR laser and a nonlinear crystal that doubles the frequency of the IR light to green light. But that process isn't perfect so some IR light still comes out of it. With the cheap green lasers (e.g. mine) there is no IR filter to block it. And my photo transistor is sensitive to IR light. But in the end I changed to a red laser when I saw that there was *a lot* of IR coming out of the laser and as your eyes do not react to it, that can be dangerous. Luckily my photo transistor reacts just as well to red light as to IR.

Step 4: The Right Frequency

By tapping the glass and recording it on the oscilloscope I saw (at least) two frequencies pop up. One appeared to be about 100 Hz, which is very low and the other around 800 Hz. That one looked like the frequency I was looking for. I didn't want that 100 Hz so I made a high-pass-filter to block it (and at the same time block low frequency noise such as the 50 Hz hum of mains). I used the Filter Wizard by Analog Devices to calculate the correct values of the parts, they not only make outstanding electronic parts, they are also very helpful with their use.
( https://www.analog.com/designtools/en/filterwizard/ ) Later I realized that the 100 Hz may have been produced by the entire glass shaking on it stem because of my tapping it.

Step 5: Closing the Loop

Now tapping the wineglass gave me some nice pictures on the oscilloscope, so it was time to test with a loudspeaker. It worked instantly, the wineglass started to resonate with a frequency of 807 Hz. From there it was simple, I amplified the signal coming from the (now filtered) photo transistor and fed it to the loudspeaker.

Step 6: Conclusion

Conclusion, it is possible to make an oscillator with a wineglass instead of an RC, LC, crystal or any other "normally used frequency determining device, but it isn't easy. At least it isn't easy the way I did it. The positioning of the laser, the wineglass and the photo transistor is extremely critical, it isn't just a millimeter forward or backwards, it is less than that, as I said to my brother, the phase of the moon influences the positioning too much.

Maybe someone knows of better, less critical ways to detect the vibrations of a wineglass (and no, a microphone does NOT work) Please tell me in the comments below.