Hello,

I built a pretty simple test circuit for a new MOSFET to check its performance at high temperature. I'm continuously switching the load on the MOSFET drain on and off at a 200kHz frequency. The load is 100ohm power resistor. Everything works fine at ambient temperature, however I'm noticing the gate resistor is getting very hot after a little while running...very hot to the touch after a few minutes (I measured the case temp of the resistor with a thermal camera to be about 300°F). This is definitely indicative that there might be a problem in the circuit, but I'm not sure where.

The gate resistor itself is a power 20W resistor with operating temp of 155°C and a low self-heating coefficient, so it should not be heating up that much.

Below are my test conditions and parameters:
MOSFET: (N-channel, 200°C Tj, 1200Vdrain)
Load: 100 ohm power resistor
Vds = 60V
Gate resistor: 20 ohm, 20W, 155°C
Gate frequency: (200kHz, 50% duty cycle square wave)
Gate voltage: 16V
Gate current (measured): around 0.5A

I'm also experiencing a voltage drop of 4 volts across the gate resistor (input gate voltage is 16 volts, output 11V at the gate).

The MOSFET output looks fine on the scope without too much ringing and the MOSFET itself is cool (attached to a heat sink), but that gate resistor is extremely hot.

I'm attaching the schematic below. I'd really appreciate any hints or suggestions at what's going on...The power dissipated through this resistor should not be that high to cause this much heating.

Thank you!

 

If your zener and FET are ok, the 49.9k resistor may be a closer to a 50 ohm resistor with the voltages you state.

 

Does D2 ever conduct?

 

Sounds like the Zener voltage is too low.

 

If the T is too high may stronger blow air onto it But from more convenient techs try to realize do the whole resistor mass is worming (IR thermo-camera photo) or only the thin outer layer. Very often for RF Mosfets may happen that resistor is not qualified for RF being too much inductive and/or overtaken by Focault effect. May alternatively, measure it by means of Nano-VNA. 99.999% of all non-smd resistors are not able to work over about 1 MHz. If not that case, may divide it in multiple SMD sections with correct summ of dissip power. Generally, those resistor ALWAYS is hot. As it have to lading-up the who know 50 000 pF C(GS) for who knows 1E8 times per second with probably 25 Volts, then N=C*V^2*f=3 KILOwatts. Okay, probably Your case is bit less drastic, but anyway the power is not small.

 

I think you have a 12V Zener. The MOSFET can handle 20V so remove the Zener and see if the voltage goes to 16V.
Or maybe you have a broken Gate on the BIG MOSFET.

 

Yes...wrong P/Ns... dang it. Should have looked more carefully. Thanks for the suggestions!

 

Off-topic observation

For your security and protection of other forum members (attraction of malicious players), you shouldn’t use your email as a user name on public forums.

Since you are a new member, I recommend that you create a new account before posting again. With a new user name.

I’ve reported this to the moderators. We'll see what they recommend regarding this thread.

 

Hey @djsfantasi I appreciate it a lot. I really didn't think too much of it when creating the username...but def yes, I'm glad you caught it! Changed it already.

Thanks again