[Phys-L] Re: Foucault pendulum circuit

[John Denker <jsd@AV8N.COM>]

John SOHL wrote:

> I've started to design my own system using LEDs and photodiodes

Good plan so far.

> with a
> simple logic circuit to trigger a timer to drive a power MOSFET to turn
> on the magnet. When I built it on the bench it worked great. When I
> installed it and connected it to the actual electromagnet it no longer
> worked correctly (gets very unstable).

Welcome to the fun & exciting world of magnet drivers.

> I guessing that the larger amount
> of inductance in the coil is feeding back into my circuit.

Feeding back?  If there's any sort of feedback, I'm not the least bit
surprised that it is problematic.

> I'd rather not reinvent the wheel.

Good.

> I didn't have much luck with web searches for circuits.

Searching was the right thing to do ... but I'm not surprised that
it was unrewarding in this case.

> Does anyone have a circuit they can share?

My suggestion:  do away with the feedback entirely.  Put all the
complexity into the logic circuit.  If you need proportional
control, implement a pulse-width-modulated controller.  Otherwise
it's even simpler;  just build a bang/bang controller.

To be explicit, the analog part of the circuit is

                     /
          ----------/  -----------------------------
          |       power switching      |           |
          |       transistor           |           |
          |                            R           L
        Power                          R           L
        Supply                         R           L
          |                            |           |
          |                            |           |
          -----------fuse---------------------------

It just can't get any simpler than that.  To make it work, use the
logic circuit to operate the transistor, applying voltage to the
coil for an artfully-chosen amount of time.

When the transistor is open-circuited, the juice in the coil decays
with the obvious L-over-R time.

The fuse is in there so the thing doesn't smoke if there's a logic
failure (or whatever) causing the transistor to stay on too long.

I repeat:  Unless I greatly misunderstand the application, you don't
need a fancy feedback loop.  You don't a constant-current supply.
/Use the Force, Luke./  In this case, use the physics of the big
inductor to help you, rather than trying to fight it with a fancy
feedback loop.

   Designing a usable feedback loop would be a piece of cake for
   a skilled analog electronics engineer ... but if you don't do
   this sort of thing for a living, it might take weeks or
   months to get up to speed.  I am not offering to teach an
   analog circuit synthesis course via email.  And just finding a
   circuit diagram on the web would not suffice, since there are
   non-obvious parameters that would need to be adjusted to fit
   your situation.  Most such circuits are specialized, not
   general-purpose solutions.

   In recent decades, most things like this have been done class-D
   as discussed above.  You would need a reeeally specialized
   reason to justify doing it with an analog feedback loop.
_______________________________________________
Phys-L mailing list
Phys-L@electron.physics.buffalo.edu
https://www.physics.buffalo.edu/mailman/listinfo/phys-l