About a week ago, our furnace, a Carrier 58rav115-20 stopped heating. My wife had to turn off the power switch, to keep the induction fan from running 24/7, even with the thermostat "OFF".
When I had a chance to look at it, I found that after being powered down for several days:
1) The induction fan comes on with the power, whether the T-stat is set ON (requesting heat, or not), or OFF.
2) Error code 21 flashes for as long as you let the furnace run (hours, if you let it).
3) Error code 21 flashes, with the T-stat R wire disconnected.
4) After the initial relay 'click' at power-on, there are no other relay clicks (no attempts to enter the ignition sequence).
5) At power-on, there is immediately 2.4VAC (two-point-four, not twenty-four) to the gas valve (there are 24VAC at the transformer secondary).
6) The gas valve blue wire is not shorted to COM, or GND (I see ~0.87 MOhm)

The control board is an 'ancient' HK42FZ009 1012-940-L (which I understand has a history of problems). I bought a used board and the symptom is the same. So, I'm wondering, is that a coincidence? Or, is the problem not in these two boards? Should I keep trying boards, or look elsewhere?

Code 21 indicates a problem with the gas valve. Make sure the green wire is grounded to the furnace chassis. Do you see the igniter glowing?

7) The green wire is properly grounded to the chassis (this is a new problem, on a heater installed for >10 years).
8) There is no voltage across the ignitor
Should the ignitor come on with the power, even if the unit isn't attempting to go beyond running the induction fan (see #4 above).

It could be the relay for the combustion blower is stuck closed. Try removing the W wire in the furnace and see if the combustion blower stops.

No, the combustion blower comes on first then the pressure switch closes and then the igniter glows.

If by "combustion blower" you mean the inducer (the blower that should run first, for 90 seconds). It runs non-stop, as soon as there is power to the furnace, whether the R, or W, or both wires are removed (see also 1, 2 and 3 above).
9) The pressure switch closes a few seconds after the inducer starts and stays closed.
10) The limit switch in line with the pressure switch (orange wire) is closed.
11) the limit switches (red wire) and fusible link are closed.

Try tapping on the relay. It will be the middle relay in a group of three.

Thanks for the suggestions so far. Tried tapping all of the relays, no luck.
12) The control board recognizes continuity on the gas valve (switching OFF, or disconnecting, results in solid ON from error LED).

Since the T-shooting guide mentions that the Gas Valve Relay may be stuck, I looked at the relays with a DVM.
The relay labeled GVR has 6 contacts. Two across the coil and two independent switches.
The coil measures 267 Ohms
One switch is open (OL on the DVM)
One switch measures 1.263 MOhm
If I energize the system, the resistance changes, but remains in the MOhm range and measured VAC across the switches, is 2.5V across one and ~3V across the other, so it doesn't appear that the switch is closed (I would expect the potential to drop, near zero, if the switches were closed).
I measure 15VAC between the coil and COM. I'm not sure how the coil circuit functions, so I'm not sure if this represents 'energized', or not. Is there a test method that I could assess whether the coil is energized and not closing the switches?
Interesting note: Twice, while I was making measurements of the GVR contacts, the induction motor switched off (first time I've had that happen, in several accumulated hours of troubleshooting). Hopeful that I had wiggled a loose contact, or found a cold solder, I went through the board and all connections, wiggling and probing with a phenolic probe, but was not able to reproduce this change in operation. It does seem to be intermittent at least.

Are there tests I could test on the gas valve itself, since that is a significant part of T-shooting this fault? I've already tried tapping it, switching it on/off, unplugging, measuring for VAC (see points 6, 7, 12 above)

On a side note, noise from the induction motor has been a long-standing issue. I found that putting pressure one of the brackets that holds the motor reduced the noise considerably. It turns out the vibration isolators were not being held correctly. I bent one tab down a hair and now, if I ever get this thing running again, it will be much quieter. Furnace:1 Me:1...