Igbt tester question

Any replies?, I’m still waiting your opinion about that tester
Regards

I guess we already answered in your previous posts....

KlausST said:

I guess we already answered in your previous posts....

Click to expand...

Right, thanks

IGBTs, or for that matter transistors, diodes, have, typically, in their datasheet, test limits
and test circuits shown for working with and qualifying devices.

I do not think this simple tester represents adequate methods for testing a device.

Just look at test circuits in datasheet and see if you can, from this devices' description,
test to those vendor supplied methods.

As a test engineer early in my career we used quite capable systems to test parts,
V and I sources, pattern generators, timing settings, etc.. to do the testing under
extensive stimulation. Many parts were done under environmental stress as well.
A $24 tester is not going to do that. Test systems I am referring to typically $100K
to > $1M USD.


Regards, Dana,.

It only test one parameter, of the maximum voltage a part can withstand, not bad for me as I’m planning to use LCR-tester also beside this one
The LCR tester here at this link


If help also showing some important parameters of the part, like for example, mosfet Vt, Cg, and RDS on , also the voltage of the body doide in the mosfet,

With these two devices I think I can easily get close to if the part have problem or no,
For the test circuits at the datasheet, as I’ve looked and it’s for switching timing test, what parameters can I get from testing circuit from datasheet except switching time tr, tf, etc?

Can you show test circuit from datasheet and we discuss here how help with identifying the part good or no?

For the test circuits at the datasheet, as I’ve looked and it’s for switching timing test, what parameters can I get from testing circuit from datasheet except switching time tr, tf, etc?

Click to expand...

Let me qualify the test work I did, it was largely digital and I/O pin drive characteristics. Testing would consist
of a gross tests to weed out crap, like dead shorts or opens on the power pins. That was to minimize wasted
time when wafer sorting and marking bad die. Test then would go on to more functional testing. Like power
up part, put it into a functional state, and measure its Poperating.

A datasheet has a lot of DC parameters, in that case circuits typically simple, these days usually not
shown in datasheet. As you point out test circuits typically shown for AC stuff. Excepting higher power stuff
where DC and safe area is discussed in detail in additional docs.

A datasheet represents both tested data and process characterization data.

In high powered devices I would not be surprised if they were doing thermal tests to insure die
attach was done effectively, eg. heat up part while measuring its time constant to insure its thermal
R in spec...... We did a simple T test by biasing substrate with high current, measuring fwd
biased diode in process to get T, and then flipping back to normal biasing and doing gross test. Helped
eliminate environmental chambers.

A breakdown test would be force a current limited V generator and measure the leakage in the
path of concern. Beta would be measured by forcing a current into base and measuring the current
out of a V source at the point transistor in active region, like Vce > 2V (that varies with transistor type)
in small signal case. Vcesat for example :



Here one forces Ibase until a measured Ic occurs (Ib x 10), and Vce is measured.

You might want to visit the OEM Test equipment sites, I would posit they have training
videos one could watch to get an idea what is typically being done these days.

In the old days we would do customer specific testing, sometimes make up particular
jigs for that purpose with custom software. I would guess these days thats only offered
in outside 3'rd party houses used to screen for specific performance.

Of course testing for life support, medical, space applications much more thorough.


Regards, Dana.

For this simple tester, it might correlate with 25'C specs for Rce= Vce(sat)/Io and hFE @2V which is basically, all you need.

Next-level performance would measure Cin, Cout (pf) @ Vce=0V which affects switching times, on/off

All other tests may also indicate quality like leakage and hFE non-linearity, but bulk resistance Rce is most critical for saturated temperature rise in switches. e.g. PN2222A is 1 ohm at rated current. a 2N3055 Rce = 0.1 ohm @ 1A, 0.15 ohm AC @ 10A, 0.078 ohm DC at 9A (typ) from plot. but 2N3055 spec says...


Rce= 0.275 ohm @ 4A
= 0.3 ohm max. at 10A at Ic/Ib = 3 !

So go get the tester or better yet, fuggedabout it and get a >=65W "lab supply" digital controlled VI power supply with V,A,W >=3 digit readouts and easy tune digital CC and CV limits. This can do the same with more accuracy and also be good as a lab supply. I find this very useful for measuring low R values and testing LED power lamps.

I also used to be a Test Eng Mgr and we had a dozen different ATE and hundreds of different functional testers. The ATE guys often used very slack tolerances for even 1% resistors to show right parts installed for ICT (roughly) but functional testers did much more, e.g. Bode plots or error rate tests.