I'm trying to produce a DC isolated output voltage of approximately 2.5V at 100mA from a ~55V DC supply. The purpose is to charge individual cells in a lead-acid battery from the power of the battery as a whole (it's a 48V system, 2V cells).

I've been playing with a PoE pulse transformer (Coilcraft POE13F-25LB).
The idea is to control pulses in the primary via PWM.

I'm switching one side of the primary to GND via an N-channel MOSFTET (Fairchild FQP45N15V2) which is driven by a MOSFET driver (Microchip MCP1415).

I'm switching at 250KHz to match the design frequency of the transformer.

I'm not sure this is the correct way to drive the pulse transformer. Here's a scope grab showing the pulse from the driver and the voltage on the MOSFET side of the transformer. enter image description here The blue trace is the gate, and shows the 1% duty cycle I was using at this point. The yellow trace shows the voltage drop when the MOSFET turns on, and then when it closes the voltage climbs again and oscillates.
This second grab shows the same situation at a 40% duty cycle: enter image description here It's this oscillation when the MOSFET turns off that's causing problems. Do I need to do something else to drive the transformer?

  • 2
    \$\begingroup\$ Why do you say it is causing a problem? Please show your circuit. Do you have any snubbers? Is this with a rectifier on the secondary? \$\endgroup\$ Commented May 11, 2017 at 2:57
  • \$\begingroup\$ Schematic is here:link Yes there's a rectifier. As for snubbing, I've had a go, but I end up dumping large currents to ground and overheating Rsnub. As for it being a problem, I'm seeing large currents through the primary and the MOSFET, and it's overheating the MOSFET. That is, currents far larger than I'd be expecing for what I'm seeing at the secondary. \$\endgroup\$ Commented May 11, 2017 at 3:50
  • \$\begingroup\$ That ringing is normal and does not cause functional problems in most applications. Your gate drive/turn off on the other hand looks weak. Food for thought: electronics.stackexchange.com/questions/246301/… \$\endgroup\$
    – winny
    Commented May 11, 2017 at 6:14
  • \$\begingroup\$ You would need a reset circuit (schottky diode + RC) across primary. \$\endgroup\$ Commented May 11, 2017 at 9:03

2 Answers 2


I wouldn't use a flyback converter for that application. You would probably do better with a forward converter with a reset winding.

In general you need feedback from the output to do a flyback as the duty cycle required depends upon the load. Since you have an isolated output it is difficult to get that feedback.

Here is a paper describing such an implementation.Single Switch forward converter.

Notice that a second winding is used with a diode to return the magnetic flux to zero after each cycle (often referred to as flux reset).

You may also be able to use a flyback converter with feedback coming from a third winding (the transformer you selected has 5 windings).

On your schematic you don't show any capacitor or load after the diode on the secondary. Also you don't need a bridge rectifier - just a single diode is needed.

Here is an example using an optocoupler-coupler

Isolated flyback converter


The oscillations are because your converter is going into discontinuous mode. It needs some minimum load. Connect a resistor or something that draws about 50 mA from the secondary.

Also, is it wrong oscilloscope settings or the yellow voltage is really going to 200 V??


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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