In our current design of our single-phase NPC T-type converter, we are using four dedicated (and expensive) power supplies (actually, DC-DC converters, since we have incoming 48V, and the converters step that down to 12V). Upon running a simulation, I have found that I can use much cheaper pulse transformers... So I would be powering four MOSFET drivers in parallel from the output of the converter, and the output of these (Vb and Vs) would be wired to each of the four primary sides of the pulse transformers. The secondary sides would be wired to the V_gs of the four MOSFETs.

How would this method compare to using four dedicated power supplies, instead of one oversized one?

Thanks a lot.

EDIT: Many are asking for schematics. This is kind of difficult since I am using new simulation software (PSIM) and I haven't mastered it to the point of making easily readable schematics. I made a very simplified one to illustrate what I am trying to say. Only the gate signal for the low-side MOSFET is shown.


simulate this circuit – Schematic created using CircuitLab

EDIT#2: I am reading through a .pdf, and the following is a proposed circuit that can be used so that pulse transformers can be implemented... Would this circuit work if correct component values are chosen? enter image description here

  • 3
    \$\begingroup\$ Schematics are superior to words. Add a sketch in with your proposal. Pulse transformers are used in AC SCR circuits which latch and are turned off when the current falls to zero. It's not clear how you expect this to drive a MOSFET and how you would switch it off sharply. \$\endgroup\$ – Transistor Aug 19 '17 at 20:08
  • \$\begingroup\$ Hi, I added a schematic... You are saying this wouldn't shut off sharply? Perhaps have a small BJT circuit would allow it to shut off quicker? \$\endgroup\$ – Lerbi Aug 20 '17 at 1:57
  • \$\begingroup\$ I don't think you've thought this through. You need to switch the MOSFET quickly on to full conduction and when switching off you need to discharge the gate capacitance very quickly. How would your transformer secondary give you precise, controllable gate voltage and time? "Perhaps have a small BJT circuit ...". If by this you mean on the secondary of the transformer then you are, in effect, using the transformer as your isolated power supply. If you mean adding it to the primary then you have gained nothing. \$\endgroup\$ – Transistor Aug 20 '17 at 5:35
  • \$\begingroup\$ Thanks a lot, @Transistor. I added yet another edit... I'm only a third year engineering student, so my knowledge on this is limited, but would the circuit I proposed work? In my simulations, it's not working... well, but I'm thinking it could be fine-tuned to work. Have you ever seen/heard of a circuit such as the one I attached? Regards, Lerbi \$\endgroup\$ – Lerbi Aug 21 '17 at 17:05
  • \$\begingroup\$ This isn't really my area of expertise and I have only once designed a high-current circuit using MOSFETs. It worked very well but I made sure that when I turned the gate on I had a BJT transistor to pull it high as quickly as possible and another one to pull it low AQAP when the opto-isolated trigger turned off. Your circuit might turn on OK but Rgs will start to turn it off immediately. If this starts a gradual turn-off of your MOSFET it will dissipate high power during the transition and may burn-out. \$\endgroup\$ – Transistor Aug 21 '17 at 19:28

Without further information in your question the following may be of interest. See Multiple MOSFET based AC switching circuits on the same supply where I posted the same response but haven't heard whether it works or not.

You may be able to solve the problem, eliminate the isolated power supplies and simplify your circuit using a photovoltaic MOSFET driver.

enter image description here

Figure 1. Extract from VOM1271 datasheet.

I haven't studied this and would be interested in comments on performance of these devices.


Yes, schematics for what you are trying would help. Sounds like you need small DC converters, a watt or two. I would first suggest selecting your topology and I would suggest a small flyback. There re small flyback converter ICs that sense the output voltage through the flyback operation. Sounds like you need several outputs, those could be additional windings on the flyback coil. A flyback transformer is very different from a pulse transformer.

You could use a pulse transformer as a forward converter, a lot depends on how much power you need to transfer.

I have a lot of experience driving large AC loads with photovoltaics, gotta watch the ON times on FETs. If you look at the secondary breakdown curves you will find slow gate rise really limits current. Check out those curves. Give me a shout if you need help in that area. Questions are answered free!

  • \$\begingroup\$ Hi, I added a schematic... Do you know why this would or wouldn't work? \$\endgroup\$ – Lerbi Aug 20 '17 at 1:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.