Help with mosfets / mosfet drivers

[FvM]

Ah, good to know... but why would I want to reduce the switching speed?

The most instructive way of coming to know is to design and operate your own switcher circuits...

You'll need at least an oscilloscope to monitor real waveforms.

 

[saburo]

Ah! Thanks Alenxan_e!
That's quite in line with what I was thinking, but I guess it might be something you can find empirically, probing the signal with a scope and checking the spikes + ringing there.
In general I would expect a small resistor between 0 and 4.7 Ohm max there, so one could just do a few prototypes and see which one performs better.
Using a trimmer to test I guess would not be the same

- - - Updated - - -

Thanks all, you really gave me a lot of information, most of which I already found looking around the web, but could not connect all the dots together, now it starts making a lot more sense.

@FvM : yup, planning to build some prototypes and ... probably watch the magic smoke come out, but that's ok. Luckily mosfets to play with come in pretty cheap I do have a 100MHz DSO (which I am already using to check the PWM generation part), hopefully no magic smoke from that one
@mtwieg : 2.5MHz sounds quite impressive, I can imagine at that level anything can interfere with your signal. Currently that's way too much for me to deal with -for now-, but it sounds like a challenging task

Overall, thanks again, you helped me a lot, now I need to digest a bit all the info and start working on some simple prototype.
I will probably bother you again

 

[saburo]

Hello again,
while i am waiting for some parts (the mosfet drivers and the mosfets) to arrive i have a very simple question, just something I wanted to double check with you.

I have seen drivers can be inverting or non inverting (or two combined inverting / non inverting) which i think comes in handy with some dc/dc converters.
I ordered non inverting ones but my understanding is i could have managed easily also with inverting types.
Is it correct that by inverting my pwm signal I obtain the same thing, but the duty cycle values are complementary i.e. if I have a 10% duty cycle and I invert the signal I will obtain a 90% duty cycle (and a shift, which is not affecting me if I am using a single fet in my experiment)?
So, since i generate the pwm with a mcu, to me it will be just a matter of configuring the software accordingly, or I am missing something?

Thanks

 

[alexan_e]

If you invert the input for the same device then you get an inverted output PWM too which mean a complimentary duty cycle as you describe but if you use an inverted input PWM and change a chip to an inverted version then the result will be exactly the same.

Depending on the driving device and output device it may be more convenient to use an inverted of normal driver version.

 

[saburo]

A small update :
I finally received the TC4420 drivers I ordered and tested them.
They seem to be really easy to use and they work fine with a 3.3V pwm signal as input.

So far I tested them only on a breadboard with jumper wires going around... I know, not the best setup and in fact I can see some massive ringing which I believe it would be less on a nicely designed PCB and adding the caps as indicated in the datasheet.
Did not have them to drive the power mosfet yet as I am first making sure at least I get the driver part right first, quite impressed by those devices so far.
I did not push them more than 1.3MHz with 15-85% duty cycle so far and -without any load on their output- they seemed to do their job nicely.

 

[saburo]

It works, I built a test circuit (still on a solderless breadboard) and I can drive the IRF3710 even at decent frequencies with this setup, but now something I am not really sure is happening, maybe you can help me figure it out.

I fear that depends on my test circuit : I am plotting on my DSO the gate signal on one channel and the VDS on the other channel.

My test circuit is extremely simple : the gate driver outputs to the mosfet gate (no gate resistor for this test) and the drain of the mosfet is connected to a resistor ("pure" resistive load for now), the other end of the resistor goes to a 12V line.

I played a bit with the frequencies and it's still quite ok at decent frequencies such as 320KHz (50% dc, did not test a wide range of duty cycles yet).

Since the second channel of the DSO is connected between drain and source, the signal goes to zero when the gate is fully open.
At 320KHz I get a nice 0-12V signal.

Now, the interesting thing is that while the gate driver output clearly shows some heavy ringing, the signal I am getting between drain and source of the mosfet looks pretty clean.
Is this some kind of filtering happening because of the gate capacitance of the Moset?
Finally, the mosfet Vgs max is normally 20V, at 12V Rds is already pretty low, IF I still have a bit of overshooting coming out of the gate driver, but the value does not exceed the 20V, I suppose it should not be a real issue, or am I wrong?
It does not seem to be reflected in the output of the power Mosfet.

Thanks for your comments