Permanent magnet assisted motor coil designs

[gyulasun]

Hi captainpecan,

Now you are talking about something similar to what I have been experimenting with all winter. Charging a coil between 2 mosfets, disconnecting the coil and at the same instant reconnecting it to a separate circuit before the field can collapse in order to collect the inductive discharge as a separate entity. I think this method has a lot of potential.

So far, due to my non-existant electronics skills,  I've fried a ton of both P&N mosfets so, please, if you would share any circuits for this I sure would appreciate it.

Regards

edit: I mean any circuits you use with your experiments on this thread..

Hi Cadman, 

I put in bold your sentence above I want to ask on:  have you attempted to connect a full wave diode bridge permanently across the coil the 2 MOSFETs were switching and direct the flyback spikes into either a capacitor or to your other circuit to utilize it?  I mean that in many cases no need for switching the flyback spikes by a dedicated switch which would need to be switched on at the moment the two MOSFETs are just switched off if you meant that?

See the attached drawing member citfta uploaded here https://overunity.com/19040/permanent-magnet-assisted-motor-coil-designs/msg564982/#msg564982  what I modified by adding a full wave bridge + a puffer capacitor.  The DC output is quasi fully isolated from the full switching circuit, the "quasi" means the body diodes of the p and n channel MOSFETs which are in reverse direction between the input supply voltage and the input of the diode bridge so the load at the output of the diode bridge normally cannot interfere with the supply voltage. 
An important note would be that in case there is a light load (i.e. a high impedance) across the DC output of the diode bridge then the flyback pulse amplitude may go high and may approach the breakdown voltages of the p and n channel MOSFETs, and can destroy them. So you should apply a relatively "heavy" load which is able to keep the flyback peak amplitude at a voltage level safe for the breakdown voltage ratings. 
The puffer cap need to have at least 200 V DC rating or higher and uF value may range from say 47 uF to 470 uF or higher. 
Note that this solution involves rectifying any induced voltage the coil may get in a particular motor setup whenever the 2 MOSFETs are switched off and those induced voltages are also collected in the puffer capacitor (this may involve a Lenz drag!).   When this is unwanted, then indeed a separate switch is needed and controlled in the needed manner if possible at all.
Note also that the MOSFET types in the schematic have only 100 V drain-source breakdown ratings so it would be advisable to use at least 400- 500 V rated types or even higher, depending on the V_cc supply voltage too.

[captainpecan]

Yup, that's the circuit I used for inspiration to do what I am doing now. I only had 1 P channel mosfet until more come in so I could only use it for 1 coil pair. I'm missing parts to complete it for now, but using a reed works great so far. I have also adapted it to use both N channel of which I have 100's of literally. I'm playing with it, but using 2 N channel so far gives me less performance and less fly back. But if I can't get it figured out, I'll have more P channels on monday.

[Cadman]

Thanks gyulasun and captainpecan.

The circuit I used is almost the same, except I used separate transistors to trigger the mosfets at the coil instead of a single trans, and a optocoupler for the signal. P-mos on the high side and N-mos on the low side of the coil. I've used a bridge on the flyback side and also tried individual diodes at the coil ends. So far having separate triggering transistors and resistors hasn't accomplished anything note-worthy.

All the mosfets and diodes are 600V 6A or better and diodes are UF type. 5V control for the transistors and 15-18V for the mosfets. The highest spike is right at 179V. Oh, and the cap collecting the discharge is a 400V 3000uF cap scavenged from a VFD.

I guess what this means is the problem is somewhere else in my circuit. Good to know.

If it ever amounts to anything I'll open a thread for it.

Thanks again guys

Cadman

[captainpecan]

Still waiting on some more P channel mosfets, so kind of slowed down my build a bit. But decided to work on a generator side of things while I wait. I wanted to choose a design I can easily couple to my motor, and be able to use it on a small wind turbine to play with if I want to. I have many odd ball ideas I want to try that this design is a good starter for. Since I have some nice strong neo magnets 60mm by 10mm bars, I chose a dual axial flux design again. Just basically like many wind turbine home builds. Seems to be a good fit for this project.There will be 2 rotors with 12 magnets on each, and a 9 coil stator sandwiched between them. I can make it relatively thin, and useful. I am bringing all coil leads out so that I can wire it in any way I wish to test things out. Here's how far I got today.

[captainpecan]

While I am waiting on parts, I was playing with the gate resistance and watching rpm and current draw changes. While playing with the pulse alignment and stuff I started noticing something interesting.


For this set up, I am just using 1 coil pair, triggered between a P and N mosfet. I posted the circuit above that I drew. Very simple. But when I move things around a little, and when i send the pulse, I start getting a large flyback spike BEFORE the pulse. I mean, at the exact moment the pulse BEGINS I am getting a spike at times. Doesn't seem to be every single pulse, or my scope just isn't catching it. Possibly there is more interferance as I was seeing earlier and got help debunking it. But the way these coils are made with an internal permanent magnet, it does flip the field outward at the beginning of the pulse, and inward at the end. I have been hoping I can somehow see that flip on the scope. Is that possibly what I am seeing here? It is showing up more, the lower the resistance I put on the gates of the mosfets. That means a little more current, more rpm, but a faster switch on time I believe that could explain catching this spike. Does anyone have any other thoughts on this leading spike I am getting at times?