Sourced the replacement for the caps that are out. Before soldering them back, I took some time to do a bit more board tracing, datasheet reading, diagnosing, etc. because I didn’t want to miss the opportunity with the caps off. I don’t want to be in a situation where, when they are back, I hope they were not (and that’s a human tendency as far as I’m aware).
The faulty section of the power supply feeds into a L4963 power regulator, which I had a close look, and which had my suspicion at one point. Its output, 5.1V, goes to all of the logic power supply pin of the logic chips or stepper motor driver chips across the board.
Checked again and again, no clear signs of short circuit measured with a DMM, either before or after this 4963. Now solder the caps back and see how it goes with a current limiting 30Ohm in series. (And a short section of aluminium flat bar is clamped behind the Darlingtons as improvised heat sink.)
As expected, fuse survived this time. No anomalies in the waveform across the added resistor, which ruled out the hypothesis of leaking rectifier diodes.
For testing and more as an indicator, the front panel is attached to the motherboard; the panel has a lcd display which previously only lit up but was blank. This time, it has stuff in display! Clearly the uC has managed to work under the less than design condition.
In the meantime, I also noticed a peculiar sizzling noise near the transformer (tiny but clearly different from the typical hum of transformer). That distracted me from the planned course of action of finding hot spot. I first thought it was something nasty going on in the 4963 chip; it was only by accident that I noticed the big guy 2200uF 50V axial cap was hot. Then it made more sense that the noise must have been from some sort of cooking in that barrel.
There may still be odds, but it’s virtually certain that this big cap is the problem.
It will take time to source its replacement (another 1000uF 16V will also be replaced). But will first use a used cylinder type cap of similar specs to see if that is the only problem, by now removing the added resistor.
This is the first time I see a big boy power filter cap that failed.