Our furnace is a Carrier Cobra GB1AAV042901ACJA made in the 39th week of 1994. About a week ago, it died without producing any diagnostic codes. The tech diagnosed the problem as the old control board and replaced it with an ICM282A he had on hand — and it fired back up again. It turns out the control board was bad, in that a contact terminal on the blower relay had blown out completely and was no longer making a connection. We celebrate the return of heat and move on…
Only, to find two days later, that the furnace has stopped heating again. Now, it simply does not produce a flame, going through an infinite loop of "run inducer motor, turn igniter (hot surface) on, turn on gas valve for a moment, turn off gas valve, turn off HSI, turn off inducer, start all over again". The control board does not blink any codes while this is happening, and does not lock out. I have run the self-test on the board and the blower motor, inducer, and HSI all pass with flying colors. Our tech has conjectured it is a stuck gas valve, or a problem with the new control board. My only theory left is to close the gas shutoff, then apply 24VAC to the gas valve via an ammeter to see if it's actually pulling current/operating normally or not operable/cutting out. Is there anything else we can do to diagnose this?
Update: further diagnosis has yielded that the gas valve is audible and a faint smell of gas is present when the furnace is trying to cycle. Applying 24VAC to the gas valve M terminal with the wire to the control board disconnected and the gas shutoff off yields 310mA or so, so clearly the gas valve is fine electrically (i.e. not open circuited). Only thing left is a control board that's either incompatible or DOA, or a gas valve that opens slightly but doesn't open fully…
Update 2: yet more diagnosis (with another ICM282A installed, as well as a replacement White-Rodgers 36J24 gas valve), and we get 500mA going to the gas valve when jumping 24V to the gas valve manually (as well as ignition, as we did this with both the gas shutoff and igniter on). However, putting the ammeter in line between the gas valve and the control wire from the board yields a mere 200-250mA. WTF? Could we have a board that's driving the gas valve too weakly? (The install manual for the valve says it only draws 280mA though…)
Best Answer
Thermostats sometimes do fail...
You know how we keep saying that HVAC problems usually aren't the thermostat's fault? Well, it turns out, whatever hit the original control over the head took out the 'stat (a Honeywell CT3500A4453 from 2007) as well, but in a subtle way. It appeared to work -- it responded normally to button pushing, and was trying to command heat. However, the W wire output from the 'stat was sagging (down to 21V from the normal 24-28V), and this was the cause of the sagging voltage going out to the gas valve, as jumpering R to W via my ammeter at the 'stat end of it all caused the furnace to fire up without a problem. Replaced the thermostat with a cheap non-programmable one, and problem solved!
Further analysis of the situation showed that the control board, instead of generating its own 24VAC signal to the gas valve, was switching the 24VAC W signal from the thermostat, which explains why the thermostat's sagging W output was causing the gas valve to malfunction. Furthermore, while I could not prove it, the only likely candidate for why the thermostat was misbehaving would be a damaged relay contact on the thermostat's control relay.