High current high voltage switching with thyristor/FET combo?

[Faringdon]

Hi,
We want to do phase cutting to control water heater power from the AC mains input.
The mains switching devices will be thyristors. It must be thyristors because if there is a short
then the thyristor can break the short circuit current. A FET  (back to back FETs as its AC) cannot do this without blowing up.
Even an IGBT would struggle.

Also, just switching a FET off when its in a mains circuit, and is carrying 40A, is very dodgy...the turn-off
loss's (overlap of voltage and current), will likely blow the FET to bits.
Therefore, we must do a Thyristor circuit, but will short out the thyristor with a FET once it starts conducting. (To reduce the conduction loss)
We will keep gate current flowing into the thyristor, so that when the FET turns OFF, the thyristor  will immediately conduct
the current...then the thyristor can do the turn off "Properly".
Do you think this i a good idea?
(Its like a synchronous thyristor drive)

[Circlotron]

Just put a big relay or a contactor in parallel with the thyristor and have the relay come on a moment after the thyristor turns on, and goes off just before the thyristor switches off. If you make the delay quick enough you may not even need a heatsink for the thyristor. The contacts will last a very long time.

AC Phase control is a very bad idea in this day and age. For heater control just switch on at the AC zero crossing and switch off whenever and let the thyristor naturally commutate. Seeing you are heating water there will be a relatively long thermal time constant, so slow cycling on and off should be fine.

[Faringdon]

Thanks, its not just water, we have another secret app where we need to switch every mains half cycle. Hence Semiconductors needed.

[mtwieg]

Hi,
We want to do phase cutting to control water heater power from the AC mains input.
The mains switching devices will be thyristors. It must be thyristors because if there is a short
then the thyristor can break the short circuit current. A FET  (back to back FETs as its AC) cannot do this without blowing up.
Even an IGBT would struggle.

Also, just switching a FET off when its in a mains circuit, and is carrying 40A, is very dodgy...the turn-off
loss's (overlap of voltage and current), will likely blow the FET to bits.

I don't understand why you'd think FETs are inherently less suitable for high frequency hard switching ("high frequency" being around 100Hz in this case), given that there are lots of other applications where FETs switch tens of amps at hundreds of kHz with no issue. I'm assuming your heater is a resistive load.

Thyristors certainly have their advantages, but switching efficiency is not one of them.

[Faringdon]

I'm assuming your heater is a resistive load.

Thanks yes but the other secret app isnt resistive...its power supplies with C at the front end.
Ayk, Electric drives, at any reasonable power, are still IGBT territory...not FETs

[ahbushnell]

I'm assuming your heater is a resistive load.

Thanks yes but the other secret app isnt resistive...its power supplies with C at the front end.
Ayk, Electric drives, at any reasonable power, are still IGBT territory...not FETs

We use SiC FETs all the time at 100kHz.  Plus electric cars/trucks use both IGBT's and SiC MOSFET's. 

[Psi]

Just put a big relay or a contactor in parallel with the thyristor and have the relay come on a moment after the thyristor turns on, and goes off just before the thyristor switches off. If you make the delay quick enough you may not even need a heatsink for the thyristor. The contacts will last a very long time.

+1 for this. (thyristor or fet across relay to make/brake the connection but relay for steady state)
Relays are awesome if you don't have huge current going through them when they switch.
Just be aware that, in some cases, you may want a small current going through the relay at turn on/off just to keep the contacts clean.