Solved! Two things that had to be fixed:
1) Shorted bridge rectifier on the rectifier board was causing only a half wave to hit the lights.
This fixed a few things:
-increased brightness of all of the controlled lamps
-increased stability of all the lamps: no dimming or flicker
-no more hum from transformer. The failed diode was likely causing the transformer to short out during 1/2 of the wave, collapsing the field and making an audible vibration in the core
-no hum from audio amplifier circuit. The DC lamps were operating on a half-wave, so they were making interference that the amps were picking up
So: lots of problems solved, except no impact on the blinking lights, which actually got worse.
Tested the output from the LM353 on the solenoid/regulator board. Rock solid 5V. Tested it at test points on the MPU, Aux driver, squawk and talk, main lamp driver. Rock solid 5V. Tested it at Vdd of one of the 4514 latching decoders: 3V. Bingo! Same on all four. Turns out the trace was bad after the test point on the lamp driver board.
So the 4514 decoders *were* failing because of low voltage, and I happened to find and fix a bunch of other issues on the way.
Thanks for your help, everyone.
Thanks for being a part of this. It was a slice.
(And check out the before and after. I think it might have got hot. New one is a bit beefier.)
D3A9A24D-30BE-402E-86FA-136110C3CEE5 (resized).jpegFBC424A2-5D99-4AA9-82B2-DC4E3A01C5F3 (resized).jpeg