Thanks for the reply.
With the incomplete information I posted, that's some good advice.
However, it didn't add up to me. Replace 1 transistor in the backbox, then instantly develop multiple shorts under the PF, even though I didn't even raise the PF when fixing the CPU board? Nah, I suspected the backbox, esp. the CPU PCB.
More thorough testing showed that, whenever a switch on row 2 closed, the corresponding entire column would also close.
Hmmm, the new transistor was bad, output stuck low? Possibly. Logic probe suggested yes. (This gave me a legitimate opportunity to state boldly, "Bring in the logic probe!")
Pulled the CPU PCB, ready to swap for another transistor, and...
...heeeeey, why are these long legs still on the transistor? These things are supposed to be trimmed after soldering. Surely they'll short against the metal chassis plate right behind the PCB, and...ooooohhhhhh...
Snip, snip, fixed.
Quoted from Ashram56:
I would suspect a short, can you test continuity between each columns directly on the switch side of the connector (ie the part which is connected to the switch themselves, not the part which is connected to the backbox)
Logic would dictate that you will see a short between several signals.