[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: A s.s. head-scratcher

Original poster: Ken or Doris Herrick <kchdlh@xxxxxxxxx>

Aron, Steve (et al)-
"A resistor in backwards"-- I like that. Would that it were that simple. I've carefully checked again & for both IGBTs, +27V or so to -27 right at gate & emitter yields no budging of the collector voltages (from about 20 or so for testing purposes). And no trace of collector current. The gates appear to have opened up inside the packages, or else the transistors have failed open. The "free wheel" diodes appear OK (using the diode-check feature of my Fluke 77 DVM).
The gate waveforms do appear "perfect", i.e. quite square, so perhaps it is the gates that have opened. I'm driving them with "totem pole"-configured emitter followers with 1 ohm resistors between those circuits and the gates. Could it be that too much current is thereby delivered? I shouldn't think so.
All of my experience with MOSFET and IGBT failure (extensive!) has led me to believe that invariably they fail shorted--either gate or transistor or both. This is a new one, but then...what else should I expect in electronics troubleshooting but new problems.
I have several CM300DU-24F's that I will try next, leaving the exterior 10K resistors alone initially. They have offset terminals but I've been crafty enough to account for that in my layout. But the Gods may well still be plotting against me as they invariably have done in the past...
All of these IGBTs I bought used, so of course it may not be Gods but rather old age that has done them in. 

Any further thoughts, most appreciated.

KCH

Tesla list wrote:

Original poster: "Aron Koscho" <mailto:kc5uto@xxxxxx><kc5uto@xxxxxx>

Ken,
Are your gate waveforms too perfect? You can tell if the gate has opened with a current probe (TEK P6021 or the like) or check across the gate resistor. It is highly unlikely the part would fail open with out you noticing (BANG). Is it possible something is setup wrong regarding the measurement? I am inclined to think this is one of those things that you spend two days trouble shooting only to find there was a bad solder joint or broken wire. Or maybe you just got a resistor in backwards... 

Aron




-----Original Message-----
From: Tesla list [<mailto:tesla@xxxxxxxxxx>mailto:tesla@xxxxxxxxxx]
Sent: Wednesday, March 14, 2007 9:19 PM
To: <mailto:tesla@xxxxxxxxxx>tesla@xxxxxxxxxx
Subject: A s.s. head-scratcher
Original poster: Ken or Doris Herrick <mailto:kchdlh@xxxxxxxxx><kchdlh@xxxxxxxxx> 

Steve Ward et al-

This one stumps me but perhaps not someone else:  I'm implementing my
push-pull s.s. primary circuit, using two Powerex CM300DY-24H
IGBTs.  Previously it appeared to function but not satisfactorily so
I've redesigned and rebuilt the drivers.  Waveforms now, right at the
IGBT gates, appear perfect.  However...  The IGBTs will conduct no
current; nada, zip, neither of them.  Nothing within them is shorted,
per my ohmmeter, and the gate:emitter nodes are "open" per the ohmmeter.

In order to accommodate my revised driver circuit, which has
back:back zener protection, I had carefully removed the 10K resistors
that were protecting the gates.  Was I not careful enough, do you
suppose?  Could the gates have actually opened up within the
packages?  >Both< of them??  Not hardly, I should think: they'd be
shorted, if anything.  Does anyone have a clue, before I have to
un-mount them & jury-rig a test-setup?

Ken Herrick