MosFET getting hot in amp - with no load

[Potentiallyincorrect]

Do your fet’s have heat sinks? Could be heat can’t get out, so temp rises. Poor thermal interface on one of them????

 

[theAnonymous1]

I've tried 5R1 gate resistor for each gate. Not much difference. The only difference I could see is the smooth transitions.
Have a look at the previous pix for comparison, The deadtime didn't change I think. So I am still stuck with the same heating. Maybe I should try to connect a higher R1 for the IRS deadtime generator? As seen in the datasheet above.

I'm a forever novice with this stuff, but your second scope shot looks like shoot-through to me. Perhaps whatever is causing it is also the cause of the DC offset. So the lower FET went from turning on too slowly (generates heat) to now rising much faster but with shoot-through (also generates heat).

 

[Topierre]

Do your fet’s have heat sinks? Could be heat can’t get out, so temp rises. Poor thermal interface on one of them????

The heatsink is ok. The fets heat up with no load, no signal

 

[Topierre]

I'm a forever novice with this stuff, but your second scope shot looks like shoot-through to me. Perhaps whatever is causing it is also the cause of the DC offset. So the lower FET went from turning on too slowly (generates heat) to now rising much faster but with shoot-through (also generates heat).

Thanks for this note, I will consider it. What do you suggest to do?

 

[Potentiallyincorrect]

In class D, power dissipated in idle can be nearly as high as when driving a signal.

 

[Topierre]

Hi guys,
Coming back to you with some interesting results:
After all tries and tests, I decided to change the 2 Fets. After that everything started to work fine, no heat at all !!!!!!!!!
My deadtime is about 100ns from high to low. Transitions became much faster than before. I put back the 16 ohm original gate resistors, and still the same good results as seen in the scope, as with the 4.7 ohm ones. So, no difference between 4.7 and 16 ohm.
Thank you all, and see you....))))

 

[Topierre]

The MOSFETs may have failed partially; the gate capacitive barrier may have become slightly conductive.... this will manifest in oscillations. Check with an oscilloscope... short the inputs. The oscillations will shoot through the gate resistor towards the driver IC... which will eventually take the driver IC out.

... same problem with a 4ohm load (and shorted inputs)??

What is the impedance of your speakers?

Please see my last post...

 

[Topierre]

Lower FET is turning on too slowly? This will increase heat. Try 4R7 gate resistor like suggested earlier.

Please see my last post...

I think 100-150ns is pretty common, larger values can increase losses from the body diode conducting,Depends on the switching frequency and mosfets chosen,ones that have larger dies and built on older technology require more, while new small ones require less,yet another parameter you have to tune in your amplifier,as if there weren't enough already

Please see my last post...