Self accelerating reed switch magnet spinner.

[TinselKoala]

The switching electronics may not need to be tightly regulated, just heavily filtered. I just tested it and there's no change in the motor behaviour at speed, with a supply of 8 1/2 to 15 volts DC to the opamp/LED strobe circuitry.


And I have to report a major error in my reporting.

The rotor isn't made from a peanut butter jar lid at all.

It is actually a Folger's Instant Coffee jar lid.

I apologize if this has caused any confusion. Replicators will have to see if they can trade their leftover PB for some instant coffee.

[MileHigh]

TK:

Keep on trucking!   Again standard disclaimer, I am just going to make some comments but I am not asking you to do anything.  You just play around at your own pace and do your thing.  The comments are meant to be generic for everyone to consider.

I can see from the scope shot that you are working in attraction mode.  From your last clip we saw the drive coil and the sensor coil are roughly opposite each other.  Notice in your scope shot above that when the MOSFET shuts OFF that the change in flux from the drive coil is picked up by the sensor coil.  Obviously the motor is functioning fine and the switching looks very clean.  In a typical Bedini motor with a coaxial drive and sensor coil, we assume that switching the drive coil ON tends to induce EMF in the sensor coil to switch the drive coil OFF.  Conversely, we assume that switching the drive coil OFF will tend to induce EMF in the sensor coil to switch the drive coil ON.  Now in your case, we have the sensor coil rotated by roughly 180 degrees and on the opposite side of the rotor.  Therefore, one can assume that switching the drive coil OFF would tend to induce EMF in the sensor coil to switch the drive coil OFF.  That would tend to reenforce the process and create a "snap" action.   Note however, that it's not that simple because we see a "doublet" impulse spike induced on the sensor coil EMF.  We assume that the op-amp input is very sensitive and this will cause a tiny glitch on the op-amp output.  From what I can see it looks like there is the initial switch OFF, followed right away by a very short switch ON for a few microseconds.  That is shown in the "extra thickness" of the rising edge of the drive coil waveform, and the "doublet" impulse we see on the sensor coil waveform.

Again, I realize that you are doing your thing and there is nothing "wrong" in the last clip and the waveforms captures you posted.  You can see how there is an advantage to having the sensor coil at 90 degrees to the drive coil to greatly reduce the mutual induction between the drive coil and the sensor coil.   Also, a very very tiny cap between the sensor coil input on the TL082 and the pin 4 ground on the TL082 should help.  You would have to scope this and find the "Goldilocks" value that just filters the EMF signal from the sensor coil a tiny little bit.  Similarly, a 0.1 uF capacitor between the -ve input on the TL082 (pin 2), and the pin 4 ground should help.  This is the output from your 10-turn potentiometer.  So adding a small decoupling cap here works to ensure that your reference threshold voltage is inherently stable.  I see that you added the 220 uF and the 0.1 uF for the power for the TL082 which is great.

What you are trying to do is create a stable voltage environment for the TL082 for it's power, as well as for the two differential inputs.  The critical thing is to not over decouple the EMF coming from the sensor coil.  You just want to give it the lightest of decoupling so that you reduce or eliminate any high frequency noise on the signal.  Too much decoupling capacitance and you risk creating an LC tank circuit which you want to avoid like the plague.

Anyway, thanks again for making clips and doing screen shots.  I am amazed that you have more than 500 clips up on your YouTube channel now.  Note that you also have a "lid motor" going.  How about them apples!

MileHigh

[MileHigh]

Just a few comments about switching and coupling as more food for thought.

For starters, since we are playing with coils and differential inputs on an op-amp, there are many ways to invert the signals.  Let's assume that we are not going to include the rotor magnets as an option, they will remain fixed.

What are the options for "inverting the logic" of the motor?

I can think of:

- swap the wires on the drive coil
- swap the wires on the sensor coil
- swap the -ve and +ve inputs on the op-amp
- you could even invert the logic on the drive coil itself and tie one end to ground instead of +12 volts and change the driver circuitry

What is the point of all of this?  Let's assume that you simply can't eliminate the mutual coupling between the drive coil and the sensor coil.  However, you should be able to use that coupling to your advantage.  Note that you have to keep in mind that the op-amp inputs are very sensitive, and the amplification gain on the output from the differential inputs is typically in the millions or greater.  Any kind of differential noise on the differential inputs in an op-amp comparator configuration will make the op-amp output toggle between ground and +12 volts at the same speed as the noise.

To use the mutual coupling to your advantage you want this to happen:  When the you switch the drive coil ON, then the EMF induced in the sensor coil from the mutual coupling wants to switch the drive coil ON.  Similarly, when you switch the drive coil OFF, the EMF induced in the sensor coil from the mutual coupling tends to switch the drive coil OFF.

So as you can see, you don't necessarily have to obsess on reducing or eliminating the EMF induced in the sensor coil from the switching of the drive coil.  You don't necessarily have to obsess on getting the sensor coil at 90 degrees and the precise offset angle to reduce or eliminate the mutual coupling.

Instead, you could easily live with a small amount of mutual coupling and try to use it to your advantage.  It's arguable that if the small amount of mutual coupling reenforces the switching, then you will get the "snap" effect in both directions.  With the right configuration and the judicious application of capacitive filtering on the -ve and +ve inputs on the op-amp, you could get rock-solid clean and robust switching.  The fun part is that you have to put your thinking cap on and figure out what the logic of the coupling is doing.  You might have to invert the logic somewhere, or you might not.

Note that the op-amp switching is already very clean, so you are not looking for extra slew rate with the "snap," you don't need it.  Rather, the "snap" is there to improve the robustness and overall noise immunity of the circuit to external influences.

MileHigh

[TinselKoala]

Yes, it was always the intention to have the sense coil at about 90 degrees from the drive coil, I just didn't have the mount made for that yet. But I do now. Now I have full control over timing (by moving the sense coil around in angular position) and dwell (the setpoint control) and I can start the rotor in either direction simply by manipulating the setpoint control, no manual starting spin necessary.

Yes, I'll scatter some decoupling caps around the circuit, and there is a bit of glitchiness, perhaps due to coupling between the coils, but overall the thing is working great, and even has a fair amount of torque once it's spinning at speed. The duty cycle for best speed is about 65-70 percent ON.

I have the switching electronics running off the low side 12 volt battery of the 24 volt stack now, no regulator or external separate PS needed.

In the photo, which is a top view, the Drive coil axis is just about the same as the twisted green wires on the left. The drive coil itself is black and hard to see but it's the same one in the same place as before. The Sense coil is mounted on a bit of threaded plastic rod, which is mounted to the black popsickle stick which is pivoted on the top pivot bearing mount, so it stays concentric with the rotor as I vary the angular position (timing).

ETA: How can you tell it's operating in attraction mode? I haven't tried it in repulsion mode yet, so I don't know for certain, but it seems to me, in my morning fog, that the waveforms and the phase relationship between the sense and drain signals will still be the same.

[MileHigh]

TK:

The angular adjustment for the sensor coil looks great!

How can I tell it's operating in attraction mode from the scope shot?  In looking at the sensor coil waveform I can see the "zero cross" is the center of the steep negative slope between the positive and negative humps.  That's also top-dead-center for the rotor magnet fly-by.

In looking at the drive coil waveform I can see that the switching is ON BEFORE the "zero cross" and it switches OFF at the "zero cross."  Hence the motor is running in attraction mode.

If I saw the drive coil being switched ON at the moment of the "zero cross" and it then switched OFF a certain amount of time AFTER the "zero cross" then I would know the motor was running in repulsion mode.

You can also see the switching angle of course.  To be more precise, you can clearly see that your actual switching threshold is actually set below the "zero cross" of approximately six volts.  Your switching threshold is set to let's say rougly five volts.  i.e.; negative one volt relative to the "zero cross."  Again, this is impossible to do with a conventional Bedini motor.

MileHigh