Standard Water Heater Failure question

[GL243]

So my Atwood gas-only water heater is failing to light again on the 2018 267BHSW, and we had the same problem repaired under warranty the first year which resulted in the standard failed circuit board replacement. They of course replaced it with another POS Atwood circuit board, so, as expected, here I am again. I ordered a new Dinosaur UIB64 board as recommended by many others on this and other RV forums, so hopefully that will be the remedy.

But the question I have is this - the first couple times it was failing to light in this most recent DeJa Vu experience I, for some weird reason, popped the pressure relief valve briefly and the thing lit up and ran OK for each of those two particular trips, relighting again each time it needed to, and allowing me to switch it off at the wall switch and then switch it back on later in the day and then running and re-lighting OK, etc. etc. This last trip, however, it's failing to light at all, I can use a hand lighter to get it to "whoomp", or light and run the typical "3 seconds" after which it stops and won't relight again, which, per other threads here should mean it just needs a new board. But does anyone know why popping the pressure valve would have caused it to suddenly relight and run OK for a couple trips?

For the record, per another recent thread here, I turned off the WH bypass valves, and released the pressure valve until no water came out of it, presumably to make sure there was an air expansion gap in the heater tank, which I understand is the other thing that might have been causing the thing to decide not light, so, unless I did that wrong, that should not be my problem (right?).

Thanks for any insights. In any event, I'll have a spare circuit board on hand now, I guess . . .

[ALJO]

Important is the adjustment of the spark igniter/sensor - remember units are mass produced. Before replacing electronic boards always follow the test procedure - just google it

[GL243]

Thanks, Aljo, you're right I should have checked that. I googled but so far all I'm seeing is what appears to be a lot of other useful troubleshooting info, but none that includes that specific test. Would you happen to have a link to suggest? Thanks.

[GL243]

Update - got the new UIB 64 Dinosaur board about a week earlier than online source said (good surprise), so worked on the water heater today. Started by lightly sanding igniter rods, tried to light again, and, interestingly, got no gas flow even this time. Earlier, I would at least hear gas flow, and hear a click, but no spark, but this time, not even gas flow.

So I installed the new board per instructions, same story, no gas flow heard, would not light. Much head-scratching, on-line troubleshooting video searching, etc. . . .

Put the old control board back on, still same problem. Per one troubleshooting video, I pulled the thermal cut-off fuse and connected directly to by-pass it, and then got gas flow, but would not light, could get it to light with a hand-lighter and then it would quit after a few seconds, just like I could get it to do before when the problem started.

So . . . put the Dino board back on, it worked fine, lighting and running.

Put the thermal cut-off fuse back in place, expecting it to not work again, but it worked. Got water up to hot temperature after 5-10 minutes, so . . . problem solved, it appears (knock on wood?). Ordered a couple more thermal cut-off fuses for spares.

If anyone has any further experience or advice on this, much appreciated.

[TCNashville]

The igniter has a dual purpose. First it ignites the propane when called for when you turn on the water heater. Second after the water heater is running it senses flame and sends that message back to the board to keep running. If the igniter is faulty it won't send the flame message to the board and will shut down the gas flow quickly. This same scenario will occur if the Thermal Cutoff is damaged or not complete.

[GL243]

Thanks for the input. I did know that about the igniter also serving as flame sensor, was surprised to learn it actually relies on a small electric current through the carbon in the flame. What puzzles me at this point, however, is that the thermal cutoff appeared to be bad, but then worked when I hooked it back up. I did scrape the connections a bit so maybe that made it work as it obviously was not burned out.