Magnet Shake Generator - how to replicate in LTspice?

Question

I made a generator for my wife's 3rd grade class, since they are learning about magnets, electromagnets and electricity. It's a simple pipe with a stack of magnets inside and you shake it to produce some power. I'm sure you've seen them.

The pipe (which is just a scrap plastic tube taken from one of my kids' "bubble wands") is about 1 inch diameter and the magnets are ceramic 3/4 inch ones from "Hobby Lobby". It's a sloppy fit, but it works. The coil is 26 gauge magnet wire and I made about 1,050 turns. At the ends of the coil, I soldered 2 LEDs with opposite polarity since it's AC and I wanted to show that. It all works, but I eventually want to go further with the project.

This leads me to my actual question. I discovered LTspice and want to replicate this in the simulation, but I have no idea how to setup the voltage source. I have been out of the electronics game for a long time and I think I'm missing something obvious. It requires a ground to simulate the circuit, but I have no idea where to put it. The more I try to do things with it, the more I don't know. Would any of you be able to show a quick way to setup this circuit, with the above information? Maybe even just estimate how much power I'm generating.

Thanks in advance.

Answer 1

Am attacking a LTspice simulation, emphasizing simplicity.
Assumptions made:

• The voltage generator amplitude is proportional to speed - shake faster gets you higher voltage. Simulation is 5 Hz (5 shakes each second).

• Shaking generates a sinusoid voltage. Probably more like a triangle voltage, because this is not a rotating machine.

On to the coil. #26 wire is likely not optimum. Assuming an average coil diameter of 1.3 inches (it is no doubt a multilayer coil), with 1050 turns, resistance is in the 15 ohm ballpark. Better to use finer wire and get more turns: requires greater care in winding (fine wire tangles easily), and is more tedious without a winding machine.
Inductance of this coil might be 0.025 Henries - likely not a big factor unless you're a really fast shaker.

Be aware that LEDs give light proportional to current flow through them. So you really want to plot LED current vs. time in LTSpice transient simulation. But to get any current, the voltage generator has to exceed a LED's forward voltage of at least a volt or two. Slow shaking gets you no current, even though peak voltage is one volt. And choosing a RED LED may get you more light than a white LED, because RED LEDs start allowing current flow at a lower voltage (about 1.6V) than WHITE LEDs (about 2.5V) LTSpice has a few different LED models: try ones of different colour.
The big, white LED chosen here (NSSWS108) doesn't generate any light until the peak voltage rises above 2.5V:

Answer 2

The easiest way to do this would be to make measurements of voltage and current from the coil and then present that to LT spice as an AC voltage source or a pulsed source.

If you wanted more detail, you could run a motor simulation like this:

Source: https://www.precisionmicrodrives.com/content/ab-025-using-spice-to-model-dc-motors/

The motor coil pictured above could be the calculated inductance of the shake coil, and the back emf could be the voltage generated from the magnet moving past the coil.