Gate drive transformer - oscilations

[Mercury]

gate drive transformer

Hello,
I'm having problems with gate drive transformers. My circuit is this:

For the test signal (buck_fire_2) i used 250kHz PWM at 50% duty cycle. The output (gate side) is terible; just a sine wave with about 40V amplitude (which probably destroyed my fet). If I lower the frequency the output is square with some sine oscilations.

So, what can I do to lower the oscilations and make it work?

Best regards,

 

[hakeen]

gate driver transformer

Hi

I guess you see some ringing, is it something like below attachment? If yes, then try to increase DC coupling capacitor value, the ringing should decrease.

Hope it helps.[/img]

 

[Mercury]

help with gate driver transformer

Actualy no. I'm driving it with a gate driver at 15V. The output is a sine wave with about 40V amplitude. The square isn't getting through. It would appear that the gate capacitance is oscilating with the transformer inductance. I tried to increase the output resistor to 10E, but no luck.

Sorry, I only have an analog oscilloscope, otherwise I would post a picture.

Any ideas guys?

Added after 58 minutes:

Managed to make a picture. I increased the C to 5uF and R to 22E. Though the result is about the same. Here's the output:

(5V/div and 2us/div)

Any ideas?

 

[eem2am]

transformer gate drive circuit

yor cct is not great. see Laszlo Balogh pdf on gate driving...youll have to search google but its ace for you

 

[hakeen]

sine wave gate driver

What is the typical values of R312 and R315? I thought they are potentiometers.

 

[E-design]

transformer gate driver

Show a picture of the waveform measured on the input of the transformer. Maybe your driving impedance is not low enough.

A bit of resistive damping in parallel on the secondary winding can take care of ringing.

Some lab photo's we have on file to show effect of driving impedance and damping on output wave.

Scale was 2V/div. Scope incorrectly shows 200mV because of probe without the multiplier sensing ring was used.

 

[FvM]

gate transformer design

Apparently the transformer leak inductance is too high. You can of course use Z-diodes to keep the Vgs voltage rating. But generally, you should watch out for a lower leak inductance transformer if you want to drive the gate directly.

 

[Mercury]

gate drive transformer ideas

Thanks guys!
That leakage inductance being to high idea is a good one. I'm using a transformer with 2000uH primary inductance and 16uH leakage. I've now read that for f=300kHz the leakage should be lower that 1uH. So I think mine is way to high.

Now, for the input. I'm using the UCC37322 driver:

And the signal at it's output (and it's the same as the signal at transformers input):

(5V/div and 2us/div)

I don't like it being that wavy. The driver is strong enough; 9A rated. Should I decrease that 22E filter resistor at the power supply?

Anyways, thanks for you help!

Best regards