Ideas for electromechanical resonance?

I'm wondering if anyone has done this and how:

I want to resonate a high carbon steel bar at its fundamental frequency to produce an audio tone. The bar is 1/8" thick, 6-1/2" long and 1-1/2" wide and resonates at about 400 Hz when struck.

The bars are supported at their "nodes" and do resonate nicely.

What I'd like is a circuit to get it producing a continuous sine wave. So far I haven't had much luck getting enough feedback to produce oscillations using a simple circuit similar to a tuning fork oscillator. (think Buliva "Acutron" watch circuit but with a high current supply and a power transistor)

I want it to work with a plain steel bar - no magnets attached to the bars nothing in contact with the bars and preferably a simple one transistor low tech circuit -

Reply to
default
Loading thread data ...

It would seem that a nice big coil at an antinode, with a DC current on it or a permanent magnet providing a bias, will couple the bar's mechanical energy nicely into electrical energy -- movement of the bar through the magnetic field will generate a voltage in the coil, while changing the coil's current will induce a changing force on the bar.

With one coil you'll see some distortion as the coil voltage will depend more or less on the coil current times the bar velocity; depending on what you want to do this may or may not make any difference.

--
www.wescottdesign.com
Reply to
Tim Wescott

Two coils would be easiest:

A pickup coil operating in variable-reluctance mode. Wind the coil on a ceramic magnet, lots of turns of fine wire, and put it close to the bar.

Here's a fancy VR pickup:

formatting link

ebay?

The other coil, the driver, can be fewer turns of fatter wire, on a e-core maybe. Run it with DC bias, class A amp maybe.

You'll need a fair amount of voltage gain from pickup to driver, an opamp driving an output transistor or power fet maybe, but I doubt you'll need a lot of actual power to keep the bar singing. A single transistor amp might work, if you get all the windings and impedances just right.

Build the pickup coil first, whack the bar, and measure the signal. Then work from there.

John

Reply to
John Larkin

Good idea. There might be some homebrew guitar pickups around. Surely a steel string would be harder to amplify than a heavy metal bar?

Reply to
default

There will be longitudinal P wave modes, flex modes, and torsion modes; probably your apparatus is in flex mode (lower frequency). It is possible to use a pull-on-the-middle motor to maintain the oscillation, probably what you want to do is to shift one of the supports to slightly off a node and use it as a microphone/motor to maintain the oscillation. Push-pull drivers of the right kind are the core of pulse-width- modulation switching power supplies, so you can drive the linear motor with a (for instance) TL494 and control the amplitude of the oscillation by the pulse-width.

Reply to
whit3rd

I foresee a significant challenge in that the heavy metal bar exhibits mere microscopic movement when it vibrates. A guitar string swings through what, several millimeters over the pickup (reluctor) when you pluck it. How much will your bar move, a few microns? Try picking that up with an ordinary surplus reluctor like a guitar pickup or a crankshaft position sensor. If you roll your own like JL says, and do some serious work on the circuitry, might could do it. The sensor would have to be just about touching the bar. I hate to be a party pooper, but I think you might have to induce a magnetic field in your hunkasteel to get anywhere.

Reply to
gearhead

.

..

Hey -- don't bother with a sensor. Just drive it near the resonant frequency. Put a big solenoid under an antinode and adjust the drive frequency until it sings. Done.

Reply to
gearhead

It would be easy to get an old guitar pickup and try it. Output voltage will be - very roughly - proportional to frequency and inverse on magnetic gap. But gain is cheap, essentially free, so a guitar pickup can probably be made to work.

Hey, bounce a laser off it! Atomic force microscopes detect really tiny deflections with a pointer-quality laser bounced off a flexing cantelever into a photodiode.

Or use a microphone, one of those cheap electret cartriges.

John

Reply to
John Larkin

e.

e

...

I didn't even think of a mic, but that would be the way to go. If this was a git-r-done job, I still think it would be simpler to do a manual adjustment when you set it up and let it then work forevermore. But let me not be the one to deny an experimenter his experimenting.

Reply to
gearhead

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.