0
\$\begingroup\$

What's a good rig to test the pulse drain current of a MOSFET? I'm thinking using a gate driver very abruptly raise the \$V_{GS}\$ from 0 to the \$V_{GS\_max}\$ for the same period in time as in the datasheet. Then directly measure the current with an ammeter (have to find ammeters that go that high). Is it not good enough? Do I need to max out the \$V_{DS}\$ as well?

Please don't suggest I obtain a curve tracer, as I don't even have an oscilloscope yet.

I'm doing this test, as in the past, I ruined an SMPS by having the MOSFET burn out, confirmed by my transistor tester. So, I ran to my local electronics parts shop & though the new MOSFET tested fine with the same transistor tester, swapping it to the SMPS did not fix it. There could be other things wrong with the SMPS, but that was the last time I trusted active components from said store.

\$\endgroup\$
9
  • 3
    \$\begingroup\$ What value is a pulse test if you have no oscilloscope to measure it with? Note the transient thermal impedance specs, no power MOSFET handles short-circuit conditions for the several seconds an ammeter takes to acquire a reading. \$\endgroup\$
    – Tim Williams
    Jul 15, 2022 at 18:59
  • 1
    \$\begingroup\$ Replacing single components in an SMPS, or most electronics, without understanding the cause, is usually futile. Consider the effect of cascading failures: perhaps a capacitor failed, causing the transistor to fail, perhaps causing still other things to fail, and then finally maybe the fuse. None of these failures may be visually apparent (burned etc.). \$\endgroup\$
    – Tim Williams
    Jul 15, 2022 at 19:04
  • 1
    \$\begingroup\$ "Can I use a shunt & a voltmeter that will register the voltage drop across it fast enough?" Better known as an oscilloscope ;) \$\endgroup\$
    – Tim Williams
    Jul 15, 2022 at 19:05
  • 1
    \$\begingroup\$ Testing pulse drain current seems to work with a "welding" generator. I have used an Arduino nano to deliver "short" pulses (0.1 ms -> 1 ms, duty cycle <=0.1% ) and measure one voltage (one R ~ 0.1 Ohm for each MOSFET)-> pulse current, one BJT, and some (5) "paralleled MOSFET". Got easily 100 A with a simple power supply (3 A max) and some "capacitors". However, use an efficient heatsink with a fan or water circulation. \$\endgroup\$
    – Antonio51
    Jul 15, 2022 at 20:45
  • 1
    \$\begingroup\$ As an engineer designing with them, I have no problem with MOSFETs working worry-free, just fine, following datasheet ratings; what are you expecting to do differently? \$\endgroup\$
    – Tim Williams
    Jul 15, 2022 at 21:48

3 Answers 3

Reset to default
3
\$\begingroup\$

Overheated MOSFETs tend to fail as a short circuit, and that often takes down other components on the board. So replacing a MOSFET without checking the other stuff risks disappointment. Did you check the fuse?

Now about your question, when it's turned on, the MOSFET won't limit drain current to a maximum value you can measure. If RdsON is low enough, it will be as much current as the power supply can provide.

They do have a maximum allowed current, above which they overheat and smoke, shoot flames, or just explode, depending on how brutal the overcurrent was. So your proposed test would simply use the MOSFET to short the power supply and it would be destroyed.

If you have doubts about your transistor being counterfeit, you can always test RdsON with a multimeter (apply a gate voltage and use ohmmeter between source and drain). But 2-wire multimeters tend not to measure below 0.1ohm so if your MOSFET has lower RdsON, it won't tell much.

\$\endgroup\$
6
  • \$\begingroup\$ How about I use stacks of lead acid batteries as power supply? \$\endgroup\$
    – Tempus Nomen
    Jul 15, 2022 at 18:53
  • 2
    \$\begingroup\$ If you set Vds=12V and turn the mosfet on, if it has RdsON=0.1 ohm, the current will be 120A and it will dissipate 1400W. It will explode. \$\endgroup\$
    – bobflux
    Jul 15, 2022 at 18:59
  • \$\begingroup\$ Some MOSFETs are large enough to have that max pulse current, continuous even... \$\endgroup\$
    – Tempus Nomen
    Jul 15, 2022 at 19:20
  • \$\begingroup\$ How about this as a compromise... I'll aim to set it up to reach the max pulse current & time, as it in the datasheet & not willy-nilly push it. \$\endgroup\$
    – Tempus Nomen
    Jul 15, 2022 at 19:23
  • 1
    \$\begingroup\$ You mean you want to test the safe operating area? \$\endgroup\$
    – bobflux
    Jul 15, 2022 at 19:26
0
\$\begingroup\$

If you want to "test" the "pulse drain current" in the datasheet, something like this may be useful (Ipulse = ~ 250 A.)

No need for an ammeter, just measuring, with Arduino ADC, the voltage on R1.
Take care of an "incidental" over-voltage on the input.

Take care also for powers on resistors (R1 & R9.)

Simulated here Rload (R1) for 1mOhm & 10 mOhm, Duty cycle = 1%.
IRFZ44 (Pulse drain current = 160 A).

enter image description here

\$\endgroup\$
0
\$\begingroup\$

How about a really "quick and dirty" way to test a MOSFET to see if it survives a given current pulse: Charge a low ESR capacitor (C) to a voltage (V) and a limiting resistor (R) with a TC short enough to be within the SOA, and then apply a voltage to the gate to discharge the capacitor with maximum current determined by V/R. You should be able to measure the ON resistance after the test to see if it has changed (which it will, due to temperature rise), but it should return to normal quickly enough. Repeat the test a few times, and if the device still works, it is probably OK.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.