Issue: The right display is missing two segments on each digit, plus the last digit is missing totally.

As you can see, the last digit is missing (a zero) + the left two segments are missing.

Solution: Lets start with the missing segments - System 11 displays are controlled by strobes (which are all the 16 digits) and segments (eg the -- in a digit). The display tube is being strobed constantly so your eyes don't realize that the digits are not lit at the same time, but serially. For each digit being strobed, the corresponding segments are being triggered.

This problem is unusual; most typically you see segments stuck on and not stuck off. Basically segments are triggered via 4049 Hex inverter IC and an UDN7180A (these fail quite often, but usually cause the "segment stuck on" problem).

As a quick test, i swapped a know good display over from a "Game Show" and the issue remained. Therefore the display itself can be excluded as the source of the error -> we now know that the issue comes from the main CPU board or the display cable. Swapping the cables over yielded the same results - its gotta be CPU.

Using the schematics to trace back the missing segments, the culprit was most likely the 6821 PIA at U51. Using a basic DMM - red lead to ground and black lead to probe the pins - revealed two pins that had different values than the rest. So the 6821 needs to go...

When soldering in a new IC, always use quality sockets. Don't be cheap. Anyway, a new PIA is in place and voila - the missing display segments are back.

Alright, next up - the missing digit. Looking at the schematics, we are looking for last strobe, the last digit of the display. In that case the answer was quite easy - oxidation ate away one of the display connections within the glass tube.

Simple, but expensive fix - replace the display tube. These are really hard to get nowadays and expensive as hell - make sure that the tube is really to blame before doing this. Anyway - tube changed and all was well.

Issue: The General Yagov kicker is not working at all. This part adds a lot of playing fun (the kicker slams the ball back brutally to the player), so seeing it work would be nice.

Solution: At that time, Williams didnt use any TIP36c transistors to drive high power coils. They used regular TIP 122 transistors to drive whats called a "snubber board" - many variations starting with C-11232. The snubber board, basically a relay, would then energize and complete the groundpath for 50v to drive a high power solenoid.

Heres a picture of a snubber board -

Furthermore, there are no schematics of the snubber boards around in the official manuals. One gentleman from german flippermarkt.de draw one a while back -

Suspecting the Snubber board to be the most likely culprit, i swapped it with a known good one - nothing. You could hear the relay clicking, but thats about it.

Next check was to make sure that 70v are actually there. Hold one lead of the DMM to GND, the other one to +70v lug, and 70v was there. Last thing to check was the coil and the wiring to the coil itself. Getting to the coil itself is a real p* in the ass, basically all of the ramps have to go. Once there, the coil was taken out and measured. An intact coil for that location should measure around 4 ohms, the one i had measured around 3 megaohms - the coil was basically open. Replaced with a known good coil brought the kicker back to live.

Issue resolved.

Issue: On a Black Knight 2000, a display segment is always lit.

Solution: Tried the display board in another machine, which showed the same error, therefore isolating the error to the display board. Looked at the manual for the corresponding segment (see below - segment "m"). Probed the corresponding UDN7180A (U16 - 12) with a logic probe and saw the output signal was stuck for the affected pin. Probed the input pins of the UDN7180A, and they were pulsing correctly.

Replacing the affected UDN7180A solved the issue.

Issue: Everytime a switch from column 8 (grn-gry) is activated, the entire row where the switch resides in is activated as well. This causes havoc to the machine (slam tilts, false scoring...). This is most noticeable on a left flipper press, as the EOS switch causes a slam tilt, essentially quitting the game on a left flipper hit.

Solution: Checked PF for obvious shorts in the column, nothing. All diodes on switches measured ok. Checked Q46 (according to manual responsible for column 8) on the MPU, found it to be defective. Replaced, issue resolved.