Checking if used IGBT brick isn't ruptured / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Checking if used IGBT brick isn't ruptured

Move Thread

LAN_403

Tonic

Wed Mar 26 2008, 09:45PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Well, in my town there's a stock-exchange where they sell used IGBT bricks (1200V 200A are pretty common, also 400A can be founded), but I would like to hunt those that aren't ruptured. Is there a reliable way to check with DDM, or perhaps I should build a electronic circuit powered by battery and check with it?

Shaun

Wed Mar 26 2008, 10:47PM

Registered Member #690 Joined: Tue May 08 2007, 03:47AM
Location: New Jersey, USA
Posts: 616

Many times you can encounter what are known a "walking dead" semiconductors, which pass all low-stress tests such as basic switching and low voltage or current conduction, but fail spectacularly when you try to use them in a tesla coil or similar device.

Just a warning; this is why many people do their best top stay away from used semiconductors.

Tonic

Wed Mar 26 2008, 10:56PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Good to hear it, but I don't have much choice - ebay is out, my local internet auctions sell bricks for minimum 50$. Those used ones cost about 10$. I also have a variac, so the blows shouldn't be so spectacular

...

Wed Mar 26 2008, 11:36PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Actually, most IGBT bricks are going to either be dead or not, because of the large dc bus filter caps they are usually associated with. There is a picture of a failed brick in the 'post your cool pics thread' that shows an extreme case of this--if the device fails it general fails 'on' which means the energy stored in the filter caps will be dissipated in the poor switch causing a catastrophic destruction. However, it is possible for one of the IGBTs in a half/full bridge module to fail without damaging adjacent devices--in that case you can still use the working half of the module

As to testing them, you could probably rig something up with a 9v battery, flashlight bulb (I would recommend putting a little bit of a load on it so make sure you don't just have a gate/source short device) and high value resistor/switch to drive the gate. Not sure if the place will like the idea of you hooking their transistors up a mystery device, but I guess its worth a try.

Steve Conner

Thu Mar 27 2008, 10:47AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I test IGBT bricks with a MOT, variac, series resistor, and scope with a high voltage probe. The trick is to short the gate to the emitter, then prod them gently with HV between emitter and collector, until they just start to break down. As long as you're careful, they just clamp like a zener diode and no damage is done.

Then you can read off the breakdown voltage from the scope and compare it to the manufacturer's data. If it's lower than spec, the device is probably one of the "walking wounded" that were mentioned. They don't always self-destruct: some inverter topologies don't even have filter caps across the IGBTs to blow them out.

I once tested five CM600 bricks like this, that I bought from another UK coiler. Four were good, but the fifth one conducted heavily, showed crazy oscillations on the scope, and burnt the ballast resistor. This fifth one also measured some leakiness on the gate: when tested with a DMM it didn't read "infinity ohms". Make of this what you will!

steve

Steve Ward

Fri Mar 28 2008, 04:32AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

This fifth one also measured some leakiness on the gate: when tested with a DMM it didn't read "infinity ohms". Make of this what you will!

Cool! i too had a few IGBTs that measured leaky on the gates (or sometimes only 1 polarity was leaky), but they were part of a half-bridge brick, where the other one was already dead.

I really like the idea of using a HV supply to check the breakdown and leakage current, i think i will do that myself from now on.

Conundrum

Fri Mar 28 2008, 08:13PM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4062

Shaun wrote ...

Many times you can encounter what are known a "walking dead" semiconductors, which pass all low-stress tests such as basic switching and low voltage or current conduction, but fail spectacularly when you try to use them in a tesla coil or similar device.

Just a warning; this is why many people do their best top stay away from used semiconductors.

yeah I know that *now*- after trying three used-but-test-OK HOT's in a TV being repaired and had them last about 0.02 seconds each!

-A

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob