Buck Converter MOSFET Heating

Thread Starter

shinetsoman

Joined Apr 30, 2009
7
Hi Friends,

I have tried gate resistance 4.7 and 10 Ohms driving voltage 12V. This circuit working normally bellow 8Amps, Crossing 8Amps the diode D1 and the input capacitor C2 are heating up. Above 10 Amps the Mosfet , diode and input capacitor are heating up and after sometime its blow up. I have designed this for drive 20A load. buck image.jpg
 

ronsimpson

Joined Oct 7, 2019
2,986
You did not give much of a schematic.
Good you chose a fast diode. Bad you chose a 1200 volt diode, when a 100 v part will work better. At room temperature, 10A current will give 2V across the diode. That is 20 watts. As the part heats up the voltage will drop back to 1.2V and 12 watts. This part will explode with out a big heat sink!
1592830881313.png
I picked a very different diode. 10A x 0.6V = 6 watts. Needs a heat sink also!
1592831431398.png
There are other problems.
 

ronsimpson

Joined Oct 7, 2019
2,986
C2 450v cap. It is most likely not rated for 8 amps. But then I don't know much about your choice.
We don't really know if L1 can handle 9A pk. If L1 saturates the MOSFEET will get hot.
 

ronsimpson

Joined Oct 7, 2019
2,986
The mosfet looks good. 10A. 0.01ohm 0.1V 1watt. (dc losses only) The duty cycle is low so …. warm.
We know nothing about how you turn on/off the part. (gate drive) You might have a problem there. (ac losses)
 

Thread Starter

shinetsoman

Joined Apr 30, 2009
7
You did not give much of a schematic.
Good you chose a fast diode. Bad you chose a 1200 volt diode, when a 100 v part will work better. At room temperature, 10A current will give 2V across the diode. That is 20 watts. As the part heats up the voltage will drop back to 1.2V and 12 watts. This part will explode with out a big heat sink!
View attachment 210433
I picked a very different diode. 10A x 0.6V = 6 watts. Needs a heat sink also!
View attachment 210434
There are other problems.
Thank you for your valuable information
 

ronsimpson

Joined Oct 7, 2019
2,986
It took me too long to see "Isolated GND". The Gate drive looks good. Applying voltage G-S is right.
R3 10k? I think you have a different resistor because the data sheet wants 8mA pin 2 to 3.
TLP250 500nS turn on/off time. Max frequency 25khz. It is a little slow and will cause some heat.

Put a heat sink on the diode! Is the MOSFET also running hot?
C2 is hot. Do you more of C2. Two capacitors will help or get a 100V capacitor. The ESR is high on C3.

1593001277311.png
 

Thread Starter

shinetsoman

Joined Apr 30, 2009
7
It took me too long to see "Isolated GND". The Gate drive looks good. Applying voltage G-S is right.
R3 10k? I think you have a different resistor because the data sheet wants 8mA pin 2 to 3.
TLP250 500nS turn on/off time. Max frequency 25khz. It is a little slow and will cause some heat.

Put a heat sink on the diode! Is the MOSFET also running hot?
C2 is hot. Do you more of C2. Two capacitors will help or get a 100V capacitor. The ESR is high on C3.

View attachment 210611
By mistake R3 is 220E, Mosfet diode and input capacitor c2 also heating up. I will try 100V c2 and Low esr c3
 

crutschow

Joined Mar 14, 2008
34,281
Inductor current rating 20Amp
If you want 20Adc output then the inductor needs to be rated for more than that to allow for the ripple current, which will exceed 20A.
can you suggest an inductor value?
On what basis did you select 220nH, which seems way too low for a 20kHz switching frequency?

I would use at least a 100μH inductor, which should give a ripple current of about 2.4A @ 20KHz switching frequency and 20A output where the ON duty-cycle is about 14/40 = 35%.
 
I have made a video on switching regulator. you can watch this video to know more about switching regulator etc.
switching regulator part 1

switching regulator part 2

I think you should choose appropriate regulator topology first with respect to power density.
 

Thread Starter

shinetsoman

Joined Apr 30, 2009
7
If you want 20Adc output then the inductor needs to be rated for more than that to allow for the ripple current, which will exceed 20A.
On what basis did you select 220nH, which seems way too low for a 20kHz switching frequency?

I would use at least a 100μH inductor, which should give a ripple current of about 2.4A @ 20KHz switching frequency and 20A output where the ON duty-cycle is about 14/40 = 35%.
Thank you it working
 
Top