dc wave generator to drive magnet array 1

[fuelcel]

I am a mechanical/industrial designer with no real foundation for curcuit design. I have a small experimental device I'm working on for myself that requires a simple electronic driver/timer. to avoid mechanical movements I want to do the following.

I have an optical device mounted in a small frame, 1.75"x1.5". this frame is suspended by four small rubber isolaters at each corner. On each side of the frame will be mounted a small neodymian magnet. About .03 away from each magnet will be a small stationary coil. I need to send a series of dc pulses to each coil in sucesion, 1-2-3-4 at a manualy adjustable setting of between 60hz and 200hz. if seen from above the coils would fire in a circular pattern. in this way I can induce an orbital occilation into the frame. The orbit only needs to be about .01" max. Operating voltage 12v.

I've been doing some reading but I'd be months away from ever understanding how to accomplish it. It seams it would be based on a 555 timing chip?

I can fabricate if there was something to follow.

thanks for the replies.
john harvey

http://www.eti-fuelcel.com

 

[VEBill]

Since you're 'Mechanical', why not use cams or levers to generate the 0.01" motion ?

One of the problems that you might run into is that the mechanical oscillation will have a natural frequency, and your drive circuit might drift in and out of phase with that natural oscillation. If you have feedback, then you can phase lock the drive with the motion, but that complicates things.

If you insist on doing it this way, then perhaps the easiest way might be to use the parallel port on a PC (and write the SW) to provide the four signals (you'll probably need external single transistor drivers).

 

[Skysoldier]

Hi John, as you said "to avoid mechanical movements". I also assume you don't want to tie up a PC to do such a simple task. The circuit can be implemented in a variety of ways, so I will try to point you in the right direction
and "natural frequencies" can be dealt with later.
1. You can use the 555 timer to generate an adjustable clock of 60-200Hz.
2. The clock is then fed to a counter IC congfigured as a two bit, binary counter.
3. The binary counter outputs are then fed to a 2 to 4 line decoder IC, to provide four sequential outputs.
4. The four outputs will then need suitable drive circuits for each coil. (a simple transistor driver wiil work fine.
As I said, this is just a synopsis of the circuit. But it will start you in the right direction.

 

[analogkid2digitalman]

Nice basic solution Skysoldier, cool!

Wheels within wheels / In a spiral array
A pattern so grand / And complex
Time after time / We lose sight of the way
Our causes can't see / Their effects.

 

[fuelcel]

Thank you gentlemen. Yes, depending on the physics of the rubber isolater (shape/material/durometer is still negotiable) along with the mass of the optics, (low mass) it will have it's own natural frequency. this frequency what ever it ends up being could be tuned out by the isolater design, if it can't take out of the 60hz to 200hz range then it will just be avoided in the adjustment.

this arrangement will give me the most "solid state" approach without mechanical movements I am trying to avoid. I know I can achieve what is proposed, aqain its a matter of curcuitry.

Skysoldier, you have given a great sysnopsis. Can I $u for laying out a curcuit? The other issue which elliminates a PC is that it must be portable an fit in a confined space of 2.6"w x 1.75"h x 1.0"d. without airflow. I see this device like a speaker design in a way. power is supplied externaly. I just need a switch, and a small dial for the frequency change.

the magnets are about .2" dia x .125 thick. The coils I imaginge (which I can wind myself) would be .375" dia x .25" long with a .2" steel core.

thanks again for the direction.

jh

 

[fuelcel]

Another thought. It might be done with two out of phase signwave generators using the full wave. This way the coils in the 12:00 and 6:00 position would alternate between + and - voltage causing the frame to be pushed and pulled at the same time, the 9:00 and 3:0 positions would be 180deg out of phase so when they are operating the frame would move in an orbit exactly the way I want. The orbit doesn't have to perfectly circular by the way. This way twice the force can be introduced into the frame,

What do you think Skysoldier? I'm favoring this concept but you might see a better way. the 555 based cucuit would be kind of digital in its drive where the out of phase sign wave driver would have a more analogue force on the frame, you follow?


jh

 

[elfgriper]

Fuelcel,
For resonance problems they make small accelerometers that will measure the amplitude of the deflection. This could be used to adjust the gain of the circuit (feedback) and dampen the signal.
-elf