appiah4 wrote on 2023-07-20, 20:15:

Now I just need pointers on what to poke at with my new toy before I start desoldering the 7406..

You already drew a nice schematic - probe the 7406 inputs and outputs. To make it easier try to find some software that will drive the keyboard status LEDs, but without any input from the keyboard. On a timer or mouse click. Otherwise you will have a hard time figuring out what is sent to keyboard and what is received - since this is a bidirectional bus. If you can't find any such software I might be able to help in a few days.

First start with sanity check - probe pins 37 and 38 on the KBC. These should be toggling as the commands to drive LED on/off are sent out.
Then you move to input pins on the 7406 - pins 13 and 3, confirm they receive signals, and that the signal level looks good.
After that you need to check pin 12 - output for CLK, and pin 4 - output for DATA, and then pins 11 and 10 of the second gate that inverts the signal.

You should find something not toggling or the signal levels being wrong, not in TTL spec. Report back the results.

You can check some pins without any special software: Pin 3 and 4/11 of the 7406 are only used when sending data to the keyboard, and should idle at a constant level (pin 3 high, pin 4/11 low) all the other time. This path is only used when the KBC drives the data line. It's only bidirectional at pin 10 of the 7406. And that's exactly what I suggest as starting point: While probing pin 3, repeatedly press NumLock. While probing pin 4, again repeatedly press numlock. You should see rectangular pulses with a width of some tens of microseconds. The rising edge may be slow, the falling edge should be sharp. Pin 13 of the 7406 is also only used when the KBC drives the line, but I'm currently unsure whether the KBC drives CLK also while receiving to throttle the keyboard if required. As receiption seems to work perfectly, I suggest starting with checking the TX path for the DATA line, which is exclusively used for sending data to the keyboard.

Apologies for resurrecting this thread, but life got in the way in a major way and half a year later I got around to desoldering the 7406 and replacing it with a new 74LS06.

Result: In Adrian's words: "It freaking works!"

After blown tantalums, desoldering and replacing a completely battery eating KBC socket, repairing almost a dozen broken traces, noticing an incorrect bios and replacing it, and now desoldering and replacing the 7406, I have a working VLB 486 motherboard, ladies and gentlemen.

Many many thanks to everyone who has helped me out through the process, I really appreciate it!