SSR Warning note

[richiem]

Over the last several years I've had one PID controller failure and several SSR failures. The last SSR death could easily have burned down my house. SSR failures can be benign, but the last one was destructive, with the SSR itself burning furiously and causing a fire in my hot tub. If my wife hadn't woken up and seen the fire, or if we had been away, the house would have burned.

I was running a 13A, 1500W heater controlled by, in this case, a 75A SSR mounted on a DIN heat sink. I felt sure that everything was spec'd out with adequate safety margins. I was wrong -- and I made a couple of poor decisions in the housing and mounting of the SSR.

I recommend that anyone using SSRs mount them in such a way that if they fail destructively nothing else will catch fire.

To me this means mounting in a steel enclosure that is not in direct contact with any flammable material, and completely separate from other flammable or critical parts.

This is 20/20 hindsight. I really wish I had done these simple things. Now I'm in the market for a new hot tub....

[etrmedia]

Wow... sorry to hear about your hot tub... Glad no one was hurt and your house wasn't damaged, though...

Do you know how this happened? I designed an overtemp cutoff for my electric brewery that kills power if the SSR fails and causes a runaway condition, but it doesn't sound like that's quite what happened with your hot tub...

[richiem]

Nope, not runaway -- I had an overtemp failsafe thermostat. The SSR just failed in a very disastrous way, with the epoxy case getting so hot that it caught fire, which then set the large plastic electrical enclosure burning, which set the wood tub frame on fire.

The SSR had plenty of thermal "overhead" and a sufficiently large current rating -- it was just a bad part. I hesitate to make a generalization, but it seems to me that some of the Chinese parts are very poorly made with an absence of any meaningful quality control. So I now recommend checking the actual operation in-circuit, checking for any previously undetected issues.

Here's what I mean -- I noticed that that particular SSR seemed to be running hotter, determined by touching the heat sink, than I remembered others doing during the 20-hour long continuous-on heat-up of the tub from a cold (50deg water) start. I just didn't associate that observation with the likelihood of ultimate failure. It was an important piece of info that just didn't properly register in the old brain-pan. If I had let intuition work, I would have swapped that SSR out with the spare I have, and compared the running temp. That might have been enough to reveal the problem.