DRSSTC motors
Finn Hammer, Sun Sept 16 2007, 07:27PMAll,
If you have been following the CCPS thread, you will probably agree, that the motor of it is basically a DRSSTC motor, with 1 major difference.
The CCPS motor is driven with a 50% duty cycle at (roughly) 1/2 the resonant frequency of the LC load. The freewheeling diodes conduct 1/2 of the time, the IGBT`s conduct the other 1/2.
Since the switching elements are big, relatively slow, bricks, special attention has to be made to the recovery behavior of the freewheeling diode.
We have seen that special care has to be taken, to make the freewheeling diode stop conducting in a graceful manner, to avoid shoot trough. And that dead time is the best way to achieve this recovery.
I have been looking at the DRSSTC and cannot for the love of god see any reason for the freewheeling diodes to conduct. The situations where it could be induced to conduct, excess dead time, could be handled with a snubber which can be designed to be extremely fast.
But if the freewheeling diode starts conducting, shoot trough is almost certain, unless generous dead time is incorporated on a permanent basis.
Would it not be better to be without the freewheeling diode in DRSSTC motors?
Would it be possible to blast them into oblivion with a "short sharp shock" for example a forward current pulse from a disc shooter that would melt the body wires off the dies, without ruining the IGBT sitting next by?
Sorry for sounding blunt, but this has been puzzling me for a while.
Cheers, Finn Hammer
Re: DRSSTC motors
Marko, Sun Sept 16 2007, 07:57PM
Definitely not, just think of what happens when interrupter turns the bridge off; it goes high impedance with a cap charged to several kV across it O_o
Marko, Sun Sept 16 2007, 07:57PM
Would it not be better to be without the freewheeling diode in DRSSTC motors?
Definitely not, just think of what happens when interrupter turns the bridge off; it goes high impedance with a cap charged to several kV across it O_o
Re: DRSSTC motors
Sulaiman, Sun Sept 16 2007, 09:12PM
I think that both igbts and mosfets have reverse diode intrinsic function
so an external diode shouldn't harm, it should be better?
Sulaiman, Sun Sept 16 2007, 09:12PM
I think that both igbts and mosfets have reverse diode intrinsic function
so an external diode shouldn't harm, it should be better?
Re: DRSSTC motors
Marko, Sun Sept 16 2007, 09:19PM
It's not really so 'intristic' in mosfet, but is rather built in to circumvent parasitic NPN transistor that may cause latch-up. I mean, mosfet can be without the diode but virtually all power mosfets have it as it does no harm.
IGBT's don't have the diode and work well without it where circuit allows it. Even when built into same package it's still actually a separate die.
Marko, Sun Sept 16 2007, 09:19PM
Sulaiman wrote ...
I think that both igbts and mosfets have reverse diode intrinsic function
so an external diode shouldn't harm, it should be better?
I think that both igbts and mosfets have reverse diode intrinsic function
so an external diode shouldn't harm, it should be better?
It's not really so 'intristic' in mosfet, but is rather built in to circumvent parasitic NPN transistor that may cause latch-up. I mean, mosfet can be without the diode but virtually all power mosfets have it as it does no harm.
IGBT's don't have the diode and work well without it where circuit allows it. Even when built into same package it's still actually a separate die.
Re: DRSSTC motors
Finn Hammer, Sun Sept 16 2007, 09:28PM
Oh my god, that`s right. After the burst they come in right handy!
(Covers my face in shame)
Cheers, Finn Hammer
Finn Hammer, Sun Sept 16 2007, 09:28PM
Marko wrote ...
Definitely not, just think of what happens when interrupter turns the bridge off; it goes high impedance with a cap charged to several kV across it O_o
Would it not be better to be without the freewheeling diode in DRSSTC motors?
Definitely not, just think of what happens when interrupter turns the bridge off; it goes high impedance with a cap charged to several kV across it O_o
Oh my god, that`s right. After the burst they come in right handy!
(Covers my face in shame)
Cheers, Finn Hammer
Re: DRSSTC motors
Steve Ward, Mon Sept 17 2007, 04:44AM
The diodes have some phase angle of conduction depending on the delays within the driver circuit. In the case of my CM300 bridge running at 40khz, the diodes conduct for a few degrees on every RF cycle.
I dont worry about it much since the IGBTs dont seem to mind. And like i said in the CCPS thread, the device capacitance seemed to cause most of the voltage transients and thus decreased (in percentage) as the bus voltage increased.
Steve Ward, Mon Sept 17 2007, 04:44AM
The diodes have some phase angle of conduction depending on the delays within the driver circuit. In the case of my CM300 bridge running at 40khz, the diodes conduct for a few degrees on every RF cycle.
I dont worry about it much since the IGBTs dont seem to mind. And like i said in the CCPS thread, the device capacitance seemed to cause most of the voltage transients and thus decreased (in percentage) as the bus voltage increased.
Re: DRSSTC motors
ragnar, Mon Sept 17 2007, 11:34AM
As Marko says, with MOSFETs the parasitic BJT occurs with the body of the MOSFET appearing as the base. They short the source metallization onto the body region to (ideally) stop the BJT ever turning on. I've read this body-source connection is what actually causes the diode to 'appear'. You can 'isolate' it (like Richie does) with a schottky diode, but this has nothing to do with IGBTs, so I'll stop there.
ragnar, Mon Sept 17 2007, 11:34AM
As Marko says, with MOSFETs the parasitic BJT occurs with the body of the MOSFET appearing as the base. They short the source metallization onto the body region to (ideally) stop the BJT ever turning on. I've read this body-source connection is what actually causes the diode to 'appear'. You can 'isolate' it (like Richie does) with a schottky diode, but this has nothing to do with IGBTs, so I'll stop there.
Print this page