IGBT's...
ChrisP
Posts: 136
Graduating from FET's to IGBT's or trying to for higher power applications..
First logic would seem to dictate that since it has a gate instead of base, partially on would be a bad thing... Is this true of IGBT's?
Do these things share current nearly as well as fets do, or should I not look at parallel sets and just go with monsters for whatever I'm working on?
Testing an IGBT, if it shares some of the characteristics of a FET I guess a gain test is out of the question?
Dual package half bridge configuration, my guess so far is that I should get a somewhat consistant resistance reading with the proper polarities back to the center of the half bridge. Is that far off?
Been reading some data sheets, there doesnt seem to be a whole lot on Google thats plain and simple basic high power info so I'm hoping for forum help since I've learned so much here already.
Thanks!
Chris
·
First logic would seem to dictate that since it has a gate instead of base, partially on would be a bad thing... Is this true of IGBT's?
Do these things share current nearly as well as fets do, or should I not look at parallel sets and just go with monsters for whatever I'm working on?
Testing an IGBT, if it shares some of the characteristics of a FET I guess a gain test is out of the question?
Dual package half bridge configuration, my guess so far is that I should get a somewhat consistant resistance reading with the proper polarities back to the center of the half bridge. Is that far off?
Been reading some data sheets, there doesnt seem to be a whole lot on Google thats plain and simple basic high power info so I'm hoping for forum help since I've learned so much here already.
Thanks!
Chris
·
Comments
http://en.wikipedia.org/wiki/IGBT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think Inside the box first and if that doesn't work..
Re-arrange what's inside the box then...
Think outside the BOX!
As metron has pointed out, it is a packaged unit consisting of a MOSFET front end for ease of control and a BJT backend for brute strength and current carrying capacity.
What is your application in terms of voltage and current?
Cheers,
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tom Sisk
http://www.siskconsult.com
·
First is that I'm working with them in 300-600 volt applications but really only in a repair scenario and thats why I'm looking at different ways to check the condition. Shorted is obvious, but there has to be a better test to tell actual condition, not just that it works or doesnt.
Second because of the same repair scenario its not unusual to run across IGBT blocks with 1 or 2 half bridge sections still functional, so it would be nice to look at these units for 12-96 volt applications. High current H bridge tinkering etc.
Lastly, its a continuing education thing. In the low voltage arena 10-24 volts I've been using the IRLZ44NS's quite happily, in singles and parallel operation. Since I'm now working with IGBT's on a limited basis at work I'd like to learn more.
Specificaly just wondering about the simple things to watch out for, quirks due to the hyrbid fet/transistor combination.
The picture I'm getting so far is that an IGBT is a very high voltage or high power transistor package with a fet drive and internal bias to simplify use. Does this cause strange things to show up in general if they are used for a much lower voltage than their rating, or do they work more like a fet than a transistor with a high fixed current gain?
An example part would be a CM100TF-12H as a single leg, or a MG100J6ES50 as a six pack. Both 600 Volt pieces, one nice thing if I can use the partials in tinkering. Back EMF shouldnt be much of an issue at 24 Volts.... lol
The usual transistor failure in a drive is dead short or completely open so the above test will confirm that.
These are good units to play with and do sometimes fail with one part still good. Use them on the bench to play with, but avoid the urge to try to reuse them in the original application!
Cheers,
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tom Sisk
http://www.siskconsult.com
·
The standard test circuit I was aware of, I'm looking more for a "Where in its life cycle is it?" kind of test. Working but worn type thing...
Parallelling requires the same consideration as paralleling plain BJT's. Usually requires some playing around with emitter resistance and making sure no transistor can heat up more than the others. Just check any text on transistors.
Life cycle; maybe with lots of time and a curve tracer you might be able to make a prediction, but these things a re pretty much on/off, work or don't work things.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tom Sisk
http://www.siskconsult.com
·
I have parralleled IGBT and they load share like a Fet one will get a little hot then the other will conduct more untill the hot one cools enough to conduct again. I have found that if you drive the IGBT over 15khz they are not as good as using a Mosfet but the IGBT will handle very high voltages and the internal diode will handle the high voltages also but an external is recomended also. Most data sheets say that a IGBT will be fully saturated with a 15vdc signal and I have used 15vdc in all of my circuits. The Powerex company http://www.pwrx.com/ has loads of info on their IGBT moduals they have some that will handle hundreds of amps even up to 2k amps. They use +15vdc to the gate to saturate the IGBT and then use -15vdc to turn it off fully but I have found that 0vdc is good enough to stop the saturation.
You should be able to find just what you are looking for.
Most of the IGBT's are manufactured here in Temecula, Ca.
Rob7
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Stephen