switching characteristics of MOSFETs, IGBTs, BJTs
Dr. Shark, Thu Feb 09 2006, 11:19AMI guess the difference between IGBTs and FETs boils down to this: FETs use only electrons to carry a current, and because electrons are mean little critters, this leads to an on-state resistance dominating the losses. However, electrons are also swift, so once the gate is turned off, the whole thing is off, for all practical purposes.
With IGBTs, there are both electrons and holes flowing through the junction, so there is a lot more stuff around to carry a current (should be only twice as much, really, anyone know the details on this?), so there is basically no resistance but just a diode drop courtesy of the additional junction. But once the gate is off, holes like to stay around, so there is this huge tail current which makes fast stwitching a pain.
So far so good. Now comes my question. Since a BJT works by the same priciple as an IGBT, it should also suffer from a slow switch-off tail. But there are lots of BJTs around which are good for 100s of MHz, so apparently they dont. Why is this, does the base current somehow fill up the holes quickly or something like that?
And, most importantly, is there any switching device that would be good for a few 100A at 3MHz, e.g. as the active device for that handheld DRSSTC I am dreaming about every night?
Re: switching characteristics of MOSFETs, IGBTs, BJTs
ragnar, Thu Feb 09 2006, 12:56PM
Well, I guess there is such a thing, Joe. It's called an array of big-arsed low capacitance MOSFETs in a class-E setup. Does that count? hehe
=)
ragnar, Thu Feb 09 2006, 12:56PM
Well, I guess there is such a thing, Joe. It's called an array of big-arsed low capacitance MOSFETs in a class-E setup. Does that count? hehe
=)
Re: switching characteristics of MOSFETs, IGBTs, BJTs
HV Enthusiast, Thu Feb 09 2006, 01:12PM
Joe,
A quick search on GOOGLE led to the following article which goes into a lot of detail of the differences between MOSFETs and IGBTs and BJTs and why affects turn-off time etc...
HV Enthusiast, Thu Feb 09 2006, 01:12PM
Joe,
A quick search on GOOGLE led to the following article which goes into a lot of detail of the differences between MOSFETs and IGBTs and BJTs and why affects turn-off time etc...
Re: switching characteristics of MOSFETs, IGBTs, BJTs
Steve Conner, Thu Feb 09 2006, 01:13PM
The difference between IGBTs and MOSFETs is "Conductivity modulation". As far as I know, in a MOSFET, the amount of electrons available to carry current is limited to a modest amount put in at the factory by doping. The designer can't just add more, or they would overwhelm the controlling effect of the gate electrode and the MOSFET would refuse to turn off.
In an IGBT, there are huge amounts of spare electrons that are normally stored out of the way and released as needed. If you like the analogy of an IGBT as a MOSFET paired with a BJT, then the "Base" current of the BJT represents the hordes of extra electrons getting pulled into action. There can be far more than "twice" as many active charge carriers than in a MOSFET.
By the same analogy, the current tail is caused by all those extra electrons taking a finite time to stop leaping around and go back to their hiding places. The current tail can be minimized by clever design, but the more you try, the more the IGBT becomes like a MOSFET in its current-carrying capabilities.
Steve Conner, Thu Feb 09 2006, 01:13PM
The difference between IGBTs and MOSFETs is "Conductivity modulation". As far as I know, in a MOSFET, the amount of electrons available to carry current is limited to a modest amount put in at the factory by doping. The designer can't just add more, or they would overwhelm the controlling effect of the gate electrode and the MOSFET would refuse to turn off.
In an IGBT, there are huge amounts of spare electrons that are normally stored out of the way and released as needed. If you like the analogy of an IGBT as a MOSFET paired with a BJT, then the "Base" current of the BJT represents the hordes of extra electrons getting pulled into action. There can be far more than "twice" as many active charge carriers than in a MOSFET.
By the same analogy, the current tail is caused by all those extra electrons taking a finite time to stop leaping around and go back to their hiding places. The current tail can be minimized by clever design, but the more you try, the more the IGBT becomes like a MOSFET in its current-carrying capabilities.
Re: switching characteristics of MOSFETs, IGBTs, BJTs
Dr. Shark, Thu Feb 09 2006, 02:04PM
OK, that makes more sense now. I was reading a couple of App notes on IGBT vs. FET that came up on google, but I guess I just didn't read the right ones
The one EVR links is much better, how the heck did did you ask google to have that turn up as the first hit?
Class E, definitely, I am actually playing with some class E stuff at the moment. But keeping up the class E conditions for a whole array sounds even more difficult! Anyway I think I might get away with a single MOSFET for my tiny coil, since the total duty cycle will have to be rather low anyway. Time will tell...
Dr. Shark, Thu Feb 09 2006, 02:04PM
OK, that makes more sense now. I was reading a couple of App notes on IGBT vs. FET that came up on google, but I guess I just didn't read the right ones
The one EVR links is much better, how the heck did did you ask google to have that turn up as the first hit? Class E, definitely, I am actually playing with some class E stuff at the moment. But keeping up the class E conditions for a whole array sounds even more difficult! Anyway I think I might get away with a single MOSFET for my tiny coil, since the total duty cycle will have to be rather low anyway. Time will tell...
Print this page