Wiring the IR2011 Correctly

[niroshans]

Hi everyone...This is my first time building an amp so any help would be really appreciated. I tried searching the forum but I couldn't find any posts to help me out.

I got the pwm signal(both inverted and noninverted and both correctly level shifted to -15V low to -10 V high) to correctly propagate to the ir2011 mosfet driver. But the waveform from the ir2011 feeding the mosfets seem to be incorrect. The ir2011 is wired as shown below in the schematic.

I am uncertain as to why the the high side mosfet waveform is the way it is. I thought the signal should be just the inverted version of the low side mosfet waveform. The attached picture shows the output from the oscilloscope. The top waveform input to the mosfet attached to the +15V rail and the bottom is attached to the -15 rail. Any help would be deeply Appreciated. Thanks.

[classd4sure]

You're starving it man!!

Ditch R20, R21, you're not using 50Vdc rails here, seriously not needed.

Then make sure you've some method to to allow pre charge of bootstrap to occure. Hint here, try tying a ~4.7K from Vs to -15.

You'll then need a delay before enabling oscillation, or do like IR and kick the low side mosfet on first.. see the circuit they use for that "start up circuit".

That should get something happening. You'll probably want to increase the bootstrap cap in value to at least 1 to 10uF according to what others had good luck with.

Try that and see what smokes,
Chris

[Pierre]

R20 is likely causing no problems, at most a 0.5V reduction in the chip's VCC, not enough to do harm. R21 is not necessary, I would leave it out.
For the bootstrap cap, I would recommend 1uF.

In order to measure the upper mosfet gate drive, tie the oscilloscope GND to source of M1 and the tip to its gate. Be sure that your oscilloscope GND is isolated from the PSU's GND or you will effectively be shorting the output to GND and... bang!

[classd4sure]

Originally posted by Pierre
R20 is likely causing no problems, at most a 0.5V reduction in the chip's VCC, not enough to do harm.

It would really do little harm to take it out too I'd imagine. In this case .. without seeing the rest of the circuit, there's no idea if he's holding inputs low for a short time before "enabling" them or just letting it rip. Since 0V is the only precharge path we may as well speed it up, the diode will take it.

[niroshans]

Thanks guys. I removed the 4.7 ohm resister, added capacitors to the +/-15 rails and changed the diode to MURS120. I originally used a fast response diode but i think this one is much better. And now it's working perfectly. I'm building on breadboard so the output is a bit noisely but I'm still amazed at how good it sounds it definately packs a pretty good puch. I also have one more question. I connected the rails to +/- 15 just to test out the circuit, should i put the 4.7 ohm resister back when i connect the circuit to +/- 50 rail. Thanks.
Niroshan

P.S. I will post the completed schematics of the entire circuit very soon.