BDC Half bridge controller mosfet keeps burning

Thread Starter


Joined Nov 29, 2023

For past some weeks, I am facing issues of high and low side mosfet keeps burning out. Above is the schematic of my half bridge controller. Before this schematic, flyback diodes D7 and D9 were not there, so I added it. But still mosfets are burning out. I am using IRF3205 now.
I am using 24 V brushed DC motor and have a LFP battery pack at around 26 V. Where I forcefully try to stop motor from rotating, power picks up for short time and High side mosfet starts to burn and flyback diode D9 also goes bad. Low side mosfet shows Drain and Source shorted. I am driving IR2104 driver through Arduino at 18kHz PWM and power of around 100W was being drawn when mosfet fails.

Motor runs for some time but when I force to stop motor, mosfet burns out.

Request help and guidance to resolve the issue.


Joined Jul 10, 2017
When you stop the motor from rotating, the load on the bridge is the very low DC resistance of the motor armature and brushes. The stall current usually very high as compared to the running current. That is what is blowing the MOSFETs.
The solution to your problem is to stop stalling the motor.
Another thing you should probably address to avoid future stall current startup problems is to use a soft start on the motor.


Joined Feb 8, 2018
Not knowing your application for this circuit, but here are a few thoughts. I had a lot of experience driving similar motors fitted to small electric vehicles many years ago.

Using a 100W motor I would have used three IRFZ44s in parallel for Q1 and Q2, although there are FET available now rated at much higher current with RDSon less than 5mohms. I assume you use Q1 for regen braking.

Dead time. Make sure Q2 turns off before Q1 turns on. I wonder if this is adjustable with IR2104.

If you are using this circuit to power an electric vehicle, it would not be possible to prevent stall as suggested above, so adequate current protection should be used. I note you have RS1, but check the LM358 is not creating too much of a delay at your operating frequency. I never used current sense resistors, but measured the voltage directly across MOSFET Q2. setting current limit using RDSon at 25C. RDSon increases with temperature, so reducing current flow as the FETs heat up.

What current rating are your Schottky diodes. D7/D9? They should be rated to carry the stall current of the motor.

What are the values of your capacitors C10/C11? I would have used something giving a total of about 10,000 uF very low ESR and connected as close to the MOSFETs as possible. Negative sides directly to GND side of the current sense.

Hope these thoughts might prove helpful.