Switching transformer circuit produces horrible distortion and ringing

Thread Starter

Vilius_Zalenas

Joined Jul 24, 2022
169
Hi,

I am an analog/power electronics beginner. I have a task to generate nearly 200V AC voltage already having a 18.4V DC rail, but I faced massive distortion and transformer ringing problem...

In the beginning, I decided to go with a 1:10 pulse transformer and extremely simple switching circuit. I am using a pulse transformer (MPN: 750032050) and N mosfet (IRF 3205). I hooked everything up on my breadboard, and turned on an oscilloscope to see what is happening. So the first picture is what I saw immediately (picture 1): both windings have extremely weird higher order harmonics coupled into the 110 kHz square signal, also there was huge voltage spike every period. After some more research, I decided to put a snubber circuit for the mosfet and I decided to disassemble my breadboard and solder everything down to an example board (to ensure perfect contacts and minimize all parasitic parameters). I provide the formulas ( which I used to determine the RC snubber parameters) in the picture. So the second time (after soldering and adding a snubber branch) I saw this (picture 3): the voltage spike is still present and the signal is nothing similar to a square form (as it should be).

Some more additional info:

1. I checked my wiring many times (it is right)

2. I am driving the mosfet from PIC32 (I am using a high current PWM pin, which is set to 110 kHz and supplying 12-13 mA of current)

3. My 18.4V rail is capable of supplying up to 1A of current. I measured the transformer current (in the operating state), it was around 370 mA, so I dont think I have a supply rail problem (the mosfet gets only warm, but not hot...)

4. The datasheet of the transformer says its operating frequency is 100 - 400 kHz, so I picked 110 kHz PWM in order to keep the frequency as low as possible to minimize the turn on and turn off mosfet delay effects to the signal.

I am not sure if I am putting enough current into my mosfet gate, I was thinking to buy a mosfet driver, but then again, 110 kHz is not an extreme frequency... (maybe you can recommend me a good driver for the given parameters). Besides the mosfet gate current, I dont have any other ideas on how to make it work properly. So I am kindly asking for any advices on how to setup that magic pulse transformer circuit (to get something similar to square signals). I thank you in advance

Vilius
 

Attachments

crutschow

Joined Mar 14, 2008
34,285
What type of AC voltage are you trying to generate?
That circuit just generates pulses as a flyback circuit. Is that what you want?
 

nsaspook

Joined Aug 27, 2009
13,086
One thing you do need is a good high frequency bypass capacitor(s) in addition to the 470uf filter cap. A 10µF and 0.1µF from DC + to DC- directly on the board. it's a pulse (type) transformer so don't expect to see nice clean waveforms with it unload.

1658682856281.png
 

Thread Starter

Vilius_Zalenas

Joined Jul 24, 2022
169
What type of AC voltage are you trying to generate?
That circuit just generates pulses as a flyback circuit. Is that what you want?
In general, yes, I dont want to overload you with the unnecessary information about my project, so my question is only about that idealy 184Vpp square signal on the secondary winding.
 

Thread Starter

Vilius_Zalenas

Joined Jul 24, 2022
169
One thing you do need is a good high frequency bypass capacitor(s) in addition to the 470uf filter cap. A 10µF and 0.1µF from DC + to DC- directly on the board. it's a pulse (type) transformer so don't expect to see nice clean waveforms with it unload.

View attachment 272127
Just like you said, I do not expect signal like from the textbooks, but I am sure that almost 700V spike should not be there, those waveforms look so bad, I think I am clearly missing something. I will put some more capacitors just like you suggested, but is there anything else to pay attention to? What are your observations on 12 mA mosfet gate current @ 110 kHz? is that enough for IRF3205? (P.S. fundamentally changing my design to your suggested LT3751 is the worst case scenario for me, as it would break the budget and PCB design at this point, I am really grateful for your work, but this time I have to make my circuit work with minor adjustments preferably)
 

Ian0

Joined Aug 7, 2020
9,671
What is the primary inductance of the transformer, and the current through the MOSFET?
You will store energy in the inductance equal to E=0.5. I^2. L
Where do you think this energy is going when you switch off the MOSFET.
A PIC32 has a maximum operating voltage of 3.6V. A IRF3205 may need 4V to switch it on. That's probably not the best piece of design work I've ever seen.
 

Thread Starter

Vilius_Zalenas

Joined Jul 24, 2022
169
What is the primary inductance of the transformer, and the current through the MOSFET?
You will store energy in the inductance equal to E=0.5. I^2. L
Where do you think this energy is going when you switch off the MOSFET.
A PIC32 has a maximum operating voltage of 3.6V. A IRF3205 may need 4V to switch it on. That's probably not the best piece of design work I've ever seen.
The current is around 370 mA, the manufacturer declares the transformer inductance to be 10 uH. So you are suggesting me some kind of mosfet driver? Can you recommend any?
 

BobTPH

Joined Jun 5, 2013
8,813
You should know what the load is before designing it.

With the secondary open, all of the energy in the inductor goes back to the primary curcuit. You want it to go to whatever load you are driving. You basically have a voltage spike generator.
 

Ian0

Joined Aug 7, 2020
9,671
The current is around 370 mA, the manufacturer declares the transformer inductance to be 10 uH. So you are suggesting me some kind of mosfet driver? Can you recommend any?
MCP1402. There are hundreds to choose from most of them work the same way - just search for "low side driver"
 
Top