After thinking about the 'blasted SSR' for a while, I started wondering if I was
being too quick to toss it out..
Looking at a few diagrams of SSRs, I got to wondering if the Zero-Voltage
detect circuit needed power?
Maybe it does need power and maybe since I was using DC, it wasn't
seeing any (DC) power flow via the load..?. (Reversed biased)..
The other question was, had only one side of the triac blown?
Maybe the other side of the triac was still good?
In either case, maybe
reversing the DC (from PV) would allow the SSR to work?
So the other night, I loosened the two output screw terminals and reversed the wires..
The next time the tank temperature got too high,
it worked!
The 800W array power was dumped..
I measured the heat of the SSR and it was only about 10 degrees above the 50 F room temp..
Current was low, about 3A and the voltage was less than 1V..
I expected more heat, but it was cool running.. So far, so good..
The 500w array input was still running 5.5A, so the tank was still getting BTUs..
So, I need to add-on another SSR, to complete the over-heat control...