MOS Power transistors and heat dissipation questions

[dynamo]

Hello all, I am using a small module like the one in this link:

https://www.aliexpress.com/item/Tri...or-Industrial-Control-5V-36V/32751789827.html

I have, for simplicity sake, a 2 Ohm coil of 20AWG wire load hooked up to the output.
I am also using the Adjustable Power Regulator in this link to vary the current output:

https://www.aliexpress.com/item/LCD...meter-Ammeter-Current-tester/32654445678.html

When I set the current output on the regulator to 5 amps, the MOS power transistors stay relatively cool.
The voltage drops to 10 Volts to regulate the current, i.e. 10V = 5A * 2 Ohms

When I set the current output on the regulator to 2.5 amps, the MOS power transistors heat up to the point where you could literally fry eggs on them.
The voltage drops to 5 Volts to regulate the current, i.e. 5V = 2.5 A * 2 Ohms

Just wondering why the voltage drop leads to such significant heat increase on the MOS transistors when, in fact, they are outputting less current?

I have tried using 12Vdc and 24Vdc power to the regulator but no change.



Thanks for any ideas anybody...

 

[(*steve*)]

Carefully measure the voltage between the gate and source terminals of the MOSFET under both conditions and let us know what they are.

 

[bob monsen]

Just wondering why the voltage drop leads to such significant heat increase on the MOS transistors when, in fact, they are outputting less current?

The important parameter isn't amps, it is power, which is amps * volts across the device. It gets confusing for SMPS, though, since the efficiency comes into it as well. Then, you need to know the average power across the device. With an SMPS, the output caps are supplying the current much of the time.

Another problem is that your 2 Ω coil of wire may be highly inductive, which may create strange feedback loops with the SMPS. So, you could have oscillation at the output, heating up the pass device. If you had an oscilloscope, you could try to narrow it down. In the meantime, perhaps try to put a big cap (maybe 1000uF) across the output to ground, in parallel with the wire, to see if that calms it down.

 

[(*steve*)]

From reading the explanation, I thought that the power was being applied externally and the mosfet was in a device which switched it on or off.

Whilst switching transients may be an issue, my first concern was that the Vgs was being supplied from an external source and was sufficient to turn the mosfet on.

If Vgs is (say) 9V when the load is operating from 10V, and 4V when it is being operated from 5V, the cause is probably due to insufficient Vgs.

Dynamo states that the device switching the mosfet is powered externally (12V or 24V) and so this shouldn't be the issue. But it is the easiest one to test first.

I don't see any evidence that the mosfet in question is being switched rapidly.

If the power supply reads 10V/5A and 5V/2.5A then the problem probably isn't due to the mosfet not being switched fully on (because you'd expect either a higher voltage for 2.5A or a lower current at 5A.

It would be useful to power the coil directly from the power supply to confirm the resistance. An reverse polarity diode across the coil would be better than a capacitor in parallel.

 

[dynamo]

Carefully measure the voltage between the gate and source terminals of the MOSFET under both conditions and let us know what they are.

OK, will do, thank you!

The important parameter isn't amps, it is power, which is amps * volts across the device. It gets confusing for SMPS, though, since the efficiency comes into it as well. Then, you need to know the average power across the device. With an SMPS, the output caps are supplying the current much of the time.

Another problem is that your 2 Ω coil of wire may be highly inductive, which may create strange feedback loops with the SMPS. So, you could have oscillation at the output, heating up the pass device. If you had an oscilloscope, you could try to narrow it down. In the meantime, perhaps try to put a big cap (maybe 1000uF) across the output to ground, in parallel with the wire, to see if that calms it down.

I have a scope and lots of caps, will try both, thanks Bob!

From reading the explanation, I thought that the power was being applied externally and the mosfet was in a device which switched it on or off.

Whilst switching transients may be an issue, my first concern was that the Vgs was being supplied from an external source and was sufficient to turn the mosfet on.

If Vgs is (say) 9V when the load is operating from 10V, and 4V when it is being operated from 5V, the cause is probably due to insufficient Vgs.

Dynamo states that the device switching the mosfet is powered externally (12V or 24V) and so this shouldn't be the issue. But it is the easiest one to test first.

I don't see any evidence that the mosfet in question is being switched rapidly.

If the power supply reads 10V/5A and 5V/2.5A then the problem probably isn't due to the mosfet not being switched fully on (because you'd expect either a higher voltage for 2.5A or a lower current at 5A.

It would be useful to power the coil directly from the power supply to confirm the resistance. An reverse polarity diode across the coil would be better than a capacitor in parallel.

The Mosfet is switching 5 seconds on, 5 seconds off, infinite loop mode.
I have some Schottkey (sp?) diodes I can add as a flyback suppressor, will try that as well.
Thank you *Steve*!