Current Switch

[markhest]

Hi-

I need to perform a current test that will apply a 40A, 6.25 ms current pulse every 5 seconds to a transformer. My first thought was to trigger a device that would deliver the current, but could be gated with the desired pulse. An SCR would be close, but it wants to stay "on" once gated. I need something that will shut off when the trigger finishes its 6.25 mS pulse. Any suggestions? Maybe there's a much easier way to do this.

Thanks!

 

[busbar]

Ther has to be other approaches, but you might search on 'GTO Thyristor.' [Meaning--'gate turn-off'.]

 

[melone]

Could you parallel up some IGBT's and simply drive the gates with a 555 timer circuit?

 

[Lewish]

What voltage are you applying to get your 40Amps?
There are power MOSFETs that can handle that current in a single device if the voltage is 60V or less. Or you can use an IGBT and get up to 600Volt capability.

No you can't drive the gates of IGBTs from a 555 timer. You don't have enough current drive capability. For that much amperage in a FET type device, you are going to need at least 250ma to get the part to saturate rapidly. Otherwise, your current pulse is going to look like a triangle waveform.

Elantec makes a nice driver, EL7155C which takes a TTL level input, but has an output supply pin which can go up to 18V, for rapid gate switching.

 

[dydt]

try to get some reading about the McMurray-Bedford inverter...2 SCR's and a commutation circuit...it might be able to give you an idea how to control the firing angle of the SCRs that will give you the required specs..

dydt

 

[IRstuff]

Let's not forget the inductive spike when you turn off your 40 A, L*di/dt. You should have some voltage limiting diodes in your circuit.

TTFN

 

[melone]

Actually LEWISH, you are wrong about the driving an IGBT. The whole advantage about IGBT's is that they have the base drive current requirements of a FET while have the power capabilities of a BJT. The whole point of an IGBT is to allow digital outputs the ability to control a high-power device. I have driven them quite well using a 555 timer circuit.

 

[Lewish]

Well, MELONE, I guess you are right. I don't know anything about driving IGBTs. I guess the experience of have catastrophic failures due to shoot-thru current when the lower IGBT failed to turn off completely, due to inadequate gate current, before the upper IGBT turned on in an inverter doesn't count for much. I also guess that having failures due to thermal overstressing, due to too long a turn-on/turn-off time caused by inadequate gate current also is not relevant.
Warm regards.

 

[melone]

It sounds like you know what you are talking about, however, I can speak from experience that you are able to drive IGBT's from a simple digital output. I guess my experience in automotive electronics dictates that if were impossible to do what I am suggesting, than there would be MILLIONS of stranded motorists all over the world. If you ever have a few minutes, try popping open your engine controller and see if you can find were your coils are connected. You might be surprised.

Also, it sounds like you are trying to operate these IGBT in a push-pull configuration. Obviously, you know that IGBT's are are succeptable to "slow" turn off times due to the gate capacitance of the front end FET. This means that they turn on much faster than they turn off unless your circuit has a good biasing resistor network on the front end.

LEWISH, I have read many of your posts and they are always very technically sound. I just saw a small inconsitency between what was posted and what I have successfully designed too and I wanted to pass along some of my experience.

 

[Lewish]

Hi MELONE. Well, noted. I am generally working with high voltages and high currents, which dictate very large gate current drive capability to ensure turn-on or turn-off at the desired rate. That rate being determined as the rate for least power dissipation while the IGBT is in the linear region. However, I also have to limit that rate somewhat to avoid generating unacceptable EMI.
I too, have driven big MOSFETs and IGBTs with digital on the gate, but it always leaves the parts at less than full saturation. The original poster was looking to switch a 40Amp pulse into some type of transformer, so I assume he wants the current to be limited by the inductance of the transformer rather than the resistance of the switching device because the switch changed states slowly. Thus, I advise using a gate driver and at least 15 volts from gate to emitter at turn-on. This will ensure maximum turn-on rate is achieved.
I wish you Best Success.