Problem with faulty treadmill PWM controller

Question

I am trying to repair a treadmill PWM controller board.

The power MOSFET (IXTH13N110) is burned out completely with all 3 pins are shorted, and as it's a bit expensive, I can't risk to replace it and test.

I removed the burned out MOSFET and checked the gate to source voltage using an oscilloscope, without any MOSFET installed. I simulated rotation sensor using a variable frequency 555 timer. when pushing the start, it counts down and then a gate drive signal is read on oscilloscope (Vgs); with frequency of 15 kHz, Vgsmax of ~11 V, and duty cycle of <1.3%. When I increase the speed using panel, the duty cycle does not change, and it just increases up to ~90% until I increase the frequency of simulated rotation sensor. then the duty cycle returns back to <1.3%. In any speed selected on panel, the duty cycle is either <1.3% or ~90%. As far as I know, by increasing the speed, the duty cycle should increase to increase the motor speed. But that is not what I get on oscilloscope.

I also replaced the original MOSFET with different one (22N60) to test the result without motor (as I am using a ~2 kVA isolation transformer to power up the board and also not mounting the original MOSFET due to the price), but it burned out just after triggering.

1. Does mounting new identical MOSFET result in correct response?

2. Is there any way to test without installing new MOSFET?

3. As the voltage at motor terminal will not exceed ~320V (At worst case), may I use a 600V MOSFET instead of 1100V?

UPDATE 1

I checked the power section tracks, and the flyback diode is also OK.

Optocoupler solderings are from another repairman.

Answer 1

in any speed selected on panel, the duty cycle is just <1.3% or ~90%. as far as i know by increasing the speed, the duty cycle should increase to increase the motor speed. but that not what i get on oscilloscope.

The controller uses tachometer frequency in a feedback loop to regulate speed. With a fixed tachometer frequency the controller will output a low or high PWM ratio as it attempts to match the motor speed to the desired treadmill speed. Your tests indicate that this feedback action is working.

With the motor disconnected the current should be very low and voltage across the FET should not exceed the supply voltage. Therefore it should not burn out unless the supply voltage exceeds the FET's rating or something is shorted (eg. flyback diode, snubber).

i replaced a mosfet with different one (22n60) to test the result with out motor, but it burned out just after triggering...

is there any way to test with out installing new mosfet?

You should replace the FET with one of similar specs. However before doing so you should check the other components in the motor circuit to make sure they are good.

Without a schematic I can't be more specific. If you don't have one then trace the circuit and understand how it works, then you can devise appropriate tests.

One trick I often use when testing mains powered equipment is to wire a 75-150 watt incandescent light bulb in series with the device under test. At low current the bulb stays cool so the filament resistance is 5~10 times less than normal and it doesn't drop much voltage. If the device tries to draw excessive current the bulb lights up and limits current.