Gate resistor

Interesting. The gate resistors we designed in as a mattter of course on MOSFETs we intended to prevent the devices from oscillating.

Presumably your original resistors were big enough to suppress sustained oscillation, but not big enough to completely suppress ringing during switch-on and switch-off. Do you hae a scope that is good enough to see ringing at a feww hundred MHz? And acess to a point on the circuit where you could see it?

-- Bill Sloman, Nijmegen

Incidentially, I don't think I've ever seen a MOSFET oscillating. I've always included the gate resistor. I should be daring some time and skip it.

Tim

-- Deep Friar: a very philosophical monk. Website:

Interesting. The gate resistors we designed in as a mattter of course on MOSFETs we intended to prevent the devices from oscillating.

Presumably your original resistors were big enough to suppress sustained oscillation, but not big enough to completely suppress ringing during switch-on and switch-off. Do you hae a scope that is good enough to see ringing at a feww hundred MHz? And acess to a point on the circuit where you could see it?

-- Bill Sloman, Nijmegen

Something to do with shoot-through? What were the resistor values?

But you are top posting again. Shows sorry lack of discipline.

John

Another way that switching losses can go up is if the MOSFET doesn't actually oscillate but that the fast edge of the switching is at a frequency way above what the inductors are good for. In that case, you are going to see a really big current spike at the turn on. I have found that this is the case often enough to check for it.

The glitch capacitively coupled to the gate can go into the driver chip and make it do funny things. I have seen a thing that looked a little like half a cycle of a sine wave on the gate just after the MOSFET has actually switched. The gate pin pulls the output of the driver below ground briefly and then ringing pulls it up. If the driver can't hold the pin near ground just after the backwards current you get this bump.

Yes. But why?

No.

Tim

Shows sorry excess of stubbornness.

John

Neet is a hair removal product.

Tim Williams a écrit :

That helps reducing D-S diode reverse recovery charge...

I have never had trouble reading a top post, and in many cases I prefer to. Do you have a brain disease or something?

Sad that an engineer can't even comprehend the engineered usefulness of appropriate top posting.

Tim

Hmm, it would, but in this case the load is resistive, returning to ~zero inbetween pulses (TL494 style waveform). That's not it :)

Tim

-- Deep Friar: a very philosophical monk. Website:

It's seldom appropriate, IMHO.

Is that happen to my head.

are

to.=20

Not only that, it seems to mess up proper quoting on your outhouse = express=20 news client.

So anyway..

Hint: gate drive transformer.

Tim

-- Deep Friar: a very philosophical monk. Website:

No, that's unrelated. Posts from Google (such as Bill's) don't seem to quote automatically. I have to add the quotation marks manually.

This seems to coincide with the header line:

so if you can disable that, please do.

Tim

-- Deep Friar: a very philosophical monk. Website:

Not only that, it seems to mess up proper quoting on your outhouse express news client.

Il Mon, 12 Apr 2010 23:59:33 -0500, Tim Williams ha scritto:

Saturated?

Nope, Bmax ~ 30mT. It's actually got "way too many" turns for this frequency, but that just means magnetizing current will be plenty low. Which also rules that out. Driver is MOSFETs anyway, so the primary waveform is essentially ideal.

Keep guessing :)

Tim