I have a system that I am designing where I am interrupting the low side of a high voltage transformer with a relay, something like this (input:
simulate this circuit – Schematic created using CircuitLab
Previously I had been using a 400V breakdown voltage MOSFET output opto-isolated solid state relay to do the switching. However, while this worked initially, it failed after 20 or so cycles. It seems that the breakdown voltage of the FET outputs had been significantly lowered, so that current now flowed even when the control to the opto was off. My belief is that the 400V breakdown voltage of the FETs was exceeded during the opening of the switch and that this caused the damage (the load is somewhat arcy).
My theory was that replacing the FET device with a relay would give it greater robustness to transients, but I am now considering whether or not to go belt-and-braces in my protection and add a gas discharge tube in parallel with the relay output in order to absorb such transients. Does this seem like a necessary or smart thing to do? Would a varistor be a better choice? I sometimes see circuits showing GDTs in series with varistors – what advantage does this bring, if any?
My thoughts were to use a GDT with a rated DC breakover voltage of
\$230V\times110\%\times\sqrt{2} = 358V\$
+whatever allowance I have to make for the GDT's tolerance. For what it's worth, the relay is rated at 250VAC, but makes no mention of its ability to withstand surges.
Best Answer
Mains relays have been used to by the million for many years. Correctly rated they are very robust, even with an inductive load like this. I have never seen a GDT used to protect a relay. Just make sure that the relays rating is well above the load current, play safe and make it 2x the load current, and use a good quality relay.