# Bidirectional AC Switch with IGBTs

I came across a "IGBT Bi-directional Switch" on the web while looking for a solution to my problem of protecting a TRIAC/SSR from a short-circuit.

Since TRIAC/SSR turn-off is not in my control, I'm quite interested in the IGBT solution. However, being relatively unfamiliar with them I wanted to ask here in case I was not fully understanding the circuit.

next to this circuit, I've also seen the following circuit which, at least, mentally I can follow and do think should work.

1. Can this switch be viably used to switch mains 220V?

2. Since they are IGBTs, if I sense the current in the mains and if it exceeds a certain threshold, can I switch these devices off immediately?

3. Can other forms of transistors be used here? If not, why so?

1. Yes it can be used to switch 220V. Viable? Well, cost and complexity could be issues for static switching applications. It's rarely done, which may be an indication.

2. Yes, you can detect and prevent destruction with a short (even if the short occurs at peak mains voltage) however you need a short-circuit-rated IGBT and you need to detect and turn off the IGBT quickly enough (and, strange as it may sound, not too quickly). Follow the recommendations of the manufacturer. These IGBTs are designed for use in VFD (Variable Frequency Drive) motor inverters.

3. I don't know offhand of other devices rated to survive a short-circuit on the mains- perhaps some beefy MOSFETs. Aside from that, an NPN transistor, an NPN darlington or a MOSFET could be used. In the latter case, in your first circuit, the body diode will share current with the external diode so you may see some additional heating in the MOSFET on top of (forward) conduction losses.

Edit:

Here is the spec from a typical short-circuit rated IGBT datasheet:

As you can see it can withstand a direct short-circuit for 10$\mu$s under realistic conditions (hot, with 300V across it).

For more on IGBT behavior under the stressful conditions of a direct short circuit, see this reference. Slowing the turn-off of the IGBT somewhat (especially during short-circuit conditions) can reduce the voltage spike due to inductance in the load circuit and the rapid change of current.

• 1) Could you elaborate on what you mean by Static Switching? 2) Are Short-Circuit IGBTs designed to sustain large avalanche energies? Why should they not be turned off too quickly? – Saad Jul 19 '14 at 18:13
• Static switching is just switching on and off, not very fast. Not PWM and not phase control. An old term. 2) not necessarily avalanche, which is caused by over-voltage rather than overcurrent. See edit above for more. – Spehro Pefhany Jul 19 '14 at 18:59