Hello everyone
It’s my first post here,
Well, I’m looking for help if circuit to make for testing mosfet or IGBT after purchase from supplier, in other words I need be able to know if the part is original or fake?
As I’m working daily on repair of Ups machines and it’s electronic circuits and they are full of mosfets and IGBT and if the part is fake , it’s a problem

I’ve searched google little, and found couple pulse test, is it ok for testing?, or you suggest better and easier circuit ?

We have variac, “ as we can get differenet voltages and rectify for dc voltage “, also function generator, 12v automotive lamp, oscilloscope and more
How can we make dynamic test for these components?

Regards

 

just buy them from a reputable supplier. It will save you the time and expense of designing and making a tester.

 

Welcome to AAC!

As I’m working daily on repair of Ups machines and it’s electronic circuits and they are full of mosfets and IGBT and if the part is fake , it’s a problem

If you're repairing equipment for pay, you should only use components from reputable sources.

How can we make dynamic test for these components?

I don't think it's practical for you to test all parameters.

I’ve searched google little, and found couple pulse test, is it ok for testing?

You need to test as the manufacturer specifies.

Using IRLZ44 as an example.

1. Gate leakage is measured at +/-16V.
2. I used a PWM to test drain current at 6A. I wasn't able to test at a higher current. I used that circuit to get an estimate of on resistance.
3. On resistance is measured using a pulse width ≤ 300us with a duty cycle of ≤ 2%. This is to prevent junction heating from affecting (increasing) resistance.
4. Threshold voltage is measured with Vds=Vgs and Id=250uA. I had read that drain voltage and short channel effects caused threshold voltage to read low. That didn't turn out to be the case. I initially measured threshold voltage by using the MOSFETs as a current sink/source. Then designed a circuit to measure with the drain and gate shorted together. Doing it correctly gave slightly lower threshold voltages, which was the opposite of what I was lead to believe.

I bought 4 types of MOSFETs from AliExpress and had to test 100% of the devices for gate leakage (because they were all packaged in plastic bags), on resistance, current up to 6A, and threshold voltage. I used 3 different circuits and just used an ammeter to measure gate leakage.

I know the dubious origins of these parts and they won't be used in any critical applications (hobby circuits only). I bought them because I was looking for affordable power MOSFETs. I'm retired and it was worth my time to pay a fraction of what the parts would have cost from a reputable source and have to test them.

 

In general, a gate leakage test setup can combine with a turn on/turn off test setup. a set voltage which should keep it off, thru a big resistor, another voltage that should switch it on, thru some resistor, and an incandescent light bulb as the load for the ON/OFF part of the test. AND YES, devices used for repair should all be checked. Getting them from a legitimate source does make passing the test a lot more likely.

 

Thanks everyone for respond to my post , and all your replies are very good and useful

Well, anyone know a tester we can buy for this purpose?, and help identify if the part is fake or original?

 

There are certain aspects regarding the methods of construction that can not be tested,
unless You go straight in with the intention of testing to destruction.

If a project is worth doing, then it's worth doing well, right from the start.
.
.
.

 

anyone know a tester we can buy for this purpose?

None that I know of. Even manufacturers don't test all parts for all parameters. Some are guaranteed by the manufacturing process, some are sampled, some are 100% tested.

identify if the part is fake or original

If you're not buying from a reputable source, you have no way of knowing if you're buying rejected parts (i.e. something someone literally picked out of the garbage and is selling), remarked/counterfeit, or genuine. You can test for the parameters that are important to you. There are some obvious signs on some counterfeit product.

I bought some IRF9540 from AliExpress. The on resistance seems to be 2-3X higher than what the datasheet specifies. Maybe I got some IRF9520's that were relabeled, maybe they were rejects, maybe they're some other P channel MOSFET. What was important to me was that they be P channel and have a high enough drain current with an on resistance that was low enough.

Since the on resistance was a bit high, I bought some IRF4905 that have a tenth of the on resistance of IRF9540.

 

This tester can not do the job I need?

https://www.amazon.com/Withstanding-Capability-Transistor-Multifunctional-500‑3500V/dp/B0CJTDD182

 

This tester can not do the job I need?

https://www.amazon.com/Withstanding-Capability-Transistor-Multifunctional-500‑3500V/dp/B0CJTDD182

That device says it's for testing "withstanding" voltage.

 

The test I suggested was to verify that a device could work in a specific application. Not every counterfeit is a failed part, and so they might work in some applications.

 

The test I suggested was to verify that a device could work in a specific application. Not every counterfeit is a failed part, and so they might work in some applications.

I don’t understand your point of view , please clarify

 

If You drastically over-specify the part from an "iffy" Supplier,
it is less likely to fail, and it will probably perform adequately.

But there's no such thing as a "too-good-to-be-true", ~$0.35 cent MOSFET.

You are probably paying more in shipping-costs than the actual component costs.

An extremely High-Power-FET is still usually less than ~$5.oo each, from a reputable supplier,
I just can't understand what the huge problem is.
I've never had to set-up a comprehensive-testing-rig for any Component from DigiKey or Mouser,
that alone is worth serious actual Money, and Time, which is also Money.
.
.
.

 

A Pass/fail test that takes much less time than replacing a defective component will be an advantage if purchasing has got a bunch of questionable components. But for products either easy to repair or not worth repairing, testing does not make sense.

 

Even if the part passes all the tests you can think of, it still may exhibit a high rate of "infant-mortality" because of poor manufacturing quality control.
As the old English saying goes You are being "penny-wise and pound-foolish".

 

Certainly there is a very big difference between an individual purchasing a part and a purchasing person deciding to buy a bunch of parts from a questionable source, which is what I was referencing. And even for a hobby-level project, it can be very handy to know that a part is not a totally failed counterfeit before installing it in some project.