sch911 wrote:
Sounds like a bad sail switch.
I love knee jerk reactions.
No, it can not possibly be a bad sail switch else it would not light at all.
Thermocouple yes, Control board, perhaps, Burner adjustment yes, Sail switch no.
The process (Ecerpted)
Stuff happens setting it up to this point.
The sail switch closes (Done)
The control board opens the gas valve, which allows gas to flow and starts generating sparks.
Many Atwood units have a single wire ignition system, In this system the wire carries about 1,000 volts AC to make sparks, and about 1/2 volt ((0.480) DC to sense flame... But we are not yet ready to discuss the problem with that idea.
The flame lights
The flame heats a metal rod, this rod is both the "Spark Plug" and the therocouple. Ideally it starts making .480 volts DC which the control board senses, When the board sees this voltage it says "AH, Working, Let's keep on heating" till it gets a cut off signal (Different cutoffs for different issues but since we are not getting there this is where I will stop)
There is an insect that likes to nest in gas jets, This can block the full flow and cause the unit to not burn hot enough to heat the rod.
The rod can get covered in carbon from a poorly adjusted burner and not heat as fast as it should.
The fuel/air mix (Air damper) may be off causing a cooler than normal flame.
In all 3 of these cases partially obstructing the exhaust may force a higher temp.. But fixing the problem is the better idea.
ALL 3 of the suspects above are cleared by ROUTINE Preventive Maintenance. Nothing more needed.
Finally: it is possible the connection between flame sensor and mother board is not that great leading to a hotter sensor being needed, or that the board is shot or nearly so and not always detecting, but these are low probability issues in this case.