Hello my Dear Friend again,

Here, you only have left to TEST the Solution I offered before, without the Single Diodes Installed per each terminal :

JUST THIS:


The LAST RESOURCE we have left is to ACCURATELY MEASURE the LEVEL OF REVERSED VOLTAGE SPIKES (without ANY ELECTRONIC BOARD CONNECTED).

JUST power a SINGLE COIL with the INPUT VOLTAGE to run Generator, THEN DISCONNECT (CAN USE A SINGLE SWITCH WHICH CAN STAND THE HV) and measure Spike, in your case would be :

1- TEST 1- 40V INPUT at 1.0 Amp or LESS.

2- TEST 2- 80V INPUT at around 2.0 Amps or LESS.

Then REPLACE ALL FET'S (PFET & NFET) BY THE REQUIRED VOLTAGE REVERSAL TOLERANCE.(IF they are available)...which I honestly, have no idea.

I have a question here...

Can you use the SAME FET's you have to EXCITE a separate block of FETS which are connected in PARALLEL (for each terminal), as the ones taking the HV Reversed Voltages?

This way not damaging the electronic board, the Opto's or the lower reverse tolerance FET's ?

In Motor Controllers, they use a lot of MOSFET in PARALLEL to STAND REVERSED VOLTAGES, as driving Higher INPUT V & A.

Wonder if we could do the same thing here?

FINALLY, and very honestly...if we can not solve this Solid State Issue...then I have no other recouse than to GO BACK to make my DUAL BRUSHED SWITCHING COMMUTATOR.

As Figuera had it originally...EXCEPT that I would be SEQUENCING a Negative and a Positive Brush at UNISON...here there are no "fragile" electronic components which could be damaged at all.

It would be very unfortunate, as I loved the System that you have designed and Built!!

It is cleaner, noiseless and more effective as frequencies driving.

Anyways, please, let me know your opinion.

In the meantime I WILL KEEP TRYING TO FIND ANOTHER SOLUTION TO THIS ISSUE !!

I HONESTLY WILL HATE TO GO BACK TO A MECHANICAL ROTARY SWITCH, as I am NOT IN THE MOOD TO DO ANYTHING OF THAT LEVEL OF WORK AT THIS TIME!!.

Regards

Ufopolitics

Hello Again,

An "in between" option we have here, before going into a  fully Mechanical Device, is to use HV RELAYS...which would be Actuated by each FET at Low Voltage...

Then we will just need 14 Relays...Seven (7) for each Negative-Positive Bank, and here there would be ZERO possibility that Reversed Voltage will go back to Electronics.

Relays have TWO separated Circuits which are COMPLETELY ISOLATED one from the other:

1- The Input Circuit at Low Voltage and High Resistance Coil [Low Amps] (electromagnet)
2- The actual HV CONTACTS Circuit...that whenever electromagnet is activated at lower voltage/lower amps, it will either close or open, (OR BOTH) the HV Contacts.

Relays can handle up to 200V Input...or more, however, running them at 110-115 Volts would be more than enough for our testing.

It is just for Testing purposes, to see if they can handle well at 50-60 Hertz Frequencies.

And so we will be able to measure Output Results...

I can do a Diagram for that...if you like.

Just like Pierre Cotnoir had...except we will not need 72, but just 14 Relays...

Regards

Ufopolitics

I have found Tito's circuit but I can't find the contex. Anyway I believe it is the real deal.

Quote from: forest on Oct 31, 2025, 03:35 PMI have found Tito's circuit but I can't find the contex. Anyway I believe it is the real deal.

Thanks Forest!

Indeed it is an interesting design!

Regards

Ufopolitics 

Hi there,

I've been following this thread with interest. I see that you've reached a bit of a stumbling block, trying to deal with reverse negative voltages with P & N FETs. I have some experience in this area and I think I have a solution so you can still achieve a solid state solution and not have to revert back to a mechanical commutator.

Have you considered using two N Channel MOSFETs in a back to back configuration, which would effectively become an AC switch that would block in both directions when off?



I've built my own MOSFET driver boards - one for single MOSFET and another with two MOSFETs for AC switching. The board uses an isolated DC/DC converter and isolated gate driver, thus removing the need to use optical tranceivers. You simply provide the PWM input via the onboard BNC connector. The PWM signal can come from a signal generator, or more commonly a microcontroller such as an Arduino. I use a Teensy 4.1 since it has quite a nice PWM block onboard and runs at 600MHz.




Do you think using AC MOSFET switches instead of the P & N channel MOSFETs that you're currently using would solve the problem you're having?

The MOSFETs I'm using (Infineon IMYH200R series) are rated to 2000V with a choice of 26, 34, 48, 89 or 123A, but actually work well up to about 2600V.I would be more than happy to supply however many you need for this project free of charge. I can also supply the Kicad or Gerber files if you want to have some manufactured by somewhere like JLCPCB.

Best regards,

Lee