Mosfet and IGBT test circuit

[Gaber Mohamed Boraey]

Hello everyone

I'm asking if you can suggest me a simple circuit that I can build for test the mosfet or igbt outside the circuit can be one circuit for both purpose or one circuit for each

I need a good way for test the part before install into machine for repair purpose

Do you have any reference for such a kind of est circuit?

Regards

Egards

 

[BradtheRad]

The insulated gate makes it necessary to perform more than one test. I was ready to suggest the gain (hfe) test on many DMM's. However that's likely to work with bjt only since it's marked BCE.

If the insulated gate ever gets perforated then it's no longer the same device. And other things can change operating characteristics.

There are simple resistance tests to try with any transistor, where you touch the bias terminal with your finger and see if the transistor conducts. However it's unreliable because the reading can sometimes stick on either end of the meter due to static charge persisting on the bias terminal.

You really need to build a jig that applies varying voltage to the mosfet, yet measure current through the body of the device. The device may turn out to be something other than what you expect. Do you have unsorted loose components? Hall sensors? JFET's? BJT's? NPN or PNP? Dual diodes?

 

[Gaber Mohamed Boraey]

You really need to build a jig that applies varying voltage to the mosfet, yet measure current through the body of the device. The device may turn out to be something other than what you expect.

How build thst?, can you show example!

 

[BradtheRad]

Example, read bias as volts, read supply as Amperes:

 

[Gaber Mohamed Boraey]

Example, read bias as volts, read supply as Amperes:

View attachment 193326

Here you use the mosfet as switch?, triggering it with resistor?, where the load?, how the ampere through the drain leg created?

 

[BradtheRad]

I increased or decreased current through the device by dialing the potentiometer from one end to the other. Bias voltage went high and low. It lets us look for any particular bias threshold which starts the device operating. We can see how smoothly the device turns On.

It's a simple test which works on Nmos or Pmos. Enhancement mode. No load. It's up to us to pinch the component between our fingers if we wish to gauge the temperature rise. Or we can make the test more complicated with a load.

 

[Gaber Mohamed Boraey]

I increased or decreased current through the device by dialing the potentiometer from one end to the other. Bias voltage went high and low. It lets us look for any particular bias threshold which starts the device operating. We can see how smoothly the device turns On.

It's a simple test which works on Nmos or Pmos. Enhancement mode. No load. It's up to us to pinch the component between our fingers if we wish to gauge the temperature rise. Or we can make the test more complicated with a load.

Okay, but I think this is not enough test for be sure that the mosfet is good and reliable under load
Right?

 

[KlausST]

Hi,

I followed a couple of your threads ... withh similar topics in testing MOSFETs and IGBTs.

I guess it´s all said a couple of times:
* such a tester is complex ... does not include only one tiny schematic .. and is done
* it needs several circuits for testing this and that
* it needs a good PCB layout and most probably Kelvin wiring for accurate measurements
* it needs controlled and pulsed currents and voltages as test signal source
* it needs S&H, ADC, microcontroller, PC interface,
* it needs microcontroller software, PC software and so on
* ....

To buy or to build such a device it needs specifiactions form your side
* WHAT parameter you want to test (multiple)
* WHAT measurement range
* WHAT is the expected accuracy
* what testing conditions (voltage source, current source...) range, accuracy, speed...

.. Thus such a device is expected to be very expensive. Way above 1k$, maybe in the 10k$ region, maybe in the 100k$ region

*****
Just look inot a MOSFET datasheet. It has many specifications. Usually for each specification there are the test conditions, maybe there even is a test circuit.
Picking just one of the many specifications ... let´s take R_DS_ON:
* it needs a kevin wiring of D and S to measure the voltage drop accurately. Maybe even an amplifier and an S&H circuit to be able to measure pulsed current
* it often is done at a specific temperature, thus you need a device (peltier) to force the temperture to the DUT
* it needs a (pulsed), adjustable voltage source for V_GS
* it needs a fast, adjustable current source for I_DS
* needs very good PCB layout to handle the high currents, not to introduce stray inductance, to condition the signals with low settling time and high accuracy
* it needs a microcontroller to manage the test timings, set the voltages and currents, read the voltage drop
* needs a display to show the result ... OR it needs an interface to the PC and a PC software
* probably I forgot some items/steps

Again: Mind that this is only to measure one single item of the datasheet specifications list.

****

If you need further assistence in this:
before posting:
* do internet investigation about the ranges, accuracy, complexitiy
* focus on one item (don´t post global questions)
* do internet research on this item
..
Then ask a detailed question

Klaus

 

[BradtheRad]

I admit I presented a severe test. Yet it tests simple operation with no load. We live in a world of uncertainty about fake components. Or broken components. There's a fine line between destructive testing and non-destructive testing. We'd like to confirm on our own. And that requires a kind of knowledge and equipment that's hard to acquire.

Note: The 10k resistor limits current in case the device is a bjt, or in case the mosfet has a perforated gate. The supply voltage should be adjustable. 9V might send several Watts through the device, suggesting the test should be brief. Testing can be destructive or non-destructive. Even overblown specs might allow low-power usage.

A step up in complexity would make the supply high voltage, and add a load. A resistive load makes sense. However a load reduces current through the mosfet. Then maybe you're testing a large assortment of parameters. Did you ask for a simple test or a complicated test?

 

[thannara123]

Make circuit to turn ON OFF the MOSFET /IGBT with a predefined variable powersupply and Gate pulse . Connect Acuurent sensing (resistor) ,
Then measure the voltage and curent accross .
match it with the data sheet .