SSRs have been a thing for a while now. OpenEVSE, in fact, uses them to drive contactors (when so configured). I've been stubbornly resisting using them to this point if for no other reason than I just thought I could do better on cost and correctness rolling my own.
I can't really objectively say that anymore. The CPC1972GSTR is a 250 mA capable 800V AC SSR that costs only 20¢ or so more than a MOC3041 and uses the exact same footprint. It's as if you took a MOC3041 and just beefed up the "driver" triac inside it so that it could drive real loads all by itself.
I'm still going to keep the MOV across the contactor coil outputs. That should keep any commutation spikes from reaching the SSR, but unless I get any big surprises putting the theory into practice, this part ought to be able to drive a contactor coil without any trouble.
Going with this part saves a ton of board space. Unfortunately, given the mechanical arrangement of the Hydra the board sizes are fixed, so that space is going to go to waste. But it'll still save me a lot of time pick-and-placing the components.
So this, plus the shift to a 5V primary power supply with a 12V boost converter on the logic board is going to be V5.0. I'll get them fabbed as soon as I test the v4.4.2 boards that are heading back down from OSHPark right now.