Midway Lazarian Maintenance 2022
02/06/2022 - Northwest Pinball and Arcade Show (2022)
Trouble began on Show Day 1 with no picture. The monitor was still energized and the game
wasn't playing blind so I swapped in the spare game PCB. That bought a few more hours but then
again, the game had no picture and wasn't playing blind. Sometimes on power cycle there was
a picture and I did catch one occasion where the game was playing, the picture would go
blank and the sound would freeze on, then it would come back on again and resume.
The game was withdrawn to the dead zone.
25/07/2022 - Game PCB #194 repair
On the bench the main PCB had a blank screen. Checking IC 11C (2621) with a scope
found:
| 1 | hi with lo pulse | 14 | hi |
| 2 | floating | 13 | lo |
| 3 | floating | 12 | clock |
| 4 | hi | 11 | clock |
| 5 | lo | 10 | clock |
| 6 | hi | 9 | clock |
| 7 | lo | 8 | lo |
IC 11C (2621) outputs didn't look correct and replacing it brought the game to life. There were
still some misplaced lines on the first sheet only. I suspected that there were
specific line effects that were confirmed by looking at the MAME driver. A quick scope
probe through IC 12H,13H,14H,15H (LS86), IC 12I,13I,14I,15I (LS85) didn't reveal anything
obviously wrong. Working back to IC 14F,15F (2114) MAME confirmed that this effect RAM
was write-only and thus couldn't be directly tested by the Arduino ICT. Using
"RAM Write All AD" to set all RAM with Data == Address showed similar flickering line
artifacts as was seen in the game. Checking IC 14F data bus outputs with a scope found
spikes & intermittent noise-like signals. IC 15F data bus outputs were all good looking
square waves. I suspected IC 14F (2114) was bad and replacing it fixed the line effects.
01/08/2022 - Game PCB #200 repair
On the bench the main PCB had a blank screen. Checking IC 11C (2621) with a scope
found that there was no CSYNC output on pin 1. Replacing IC 11C (2621) fixed the game.
For both game PCBs I was able to find older but ceramic Signetics 2621 in the hope that
they'd be more reliable than the plastic versions.
03/12/2022 - Game PCB #200 bench test
Testing the game PCB back in the cabinet found that the game wouldn't coin up. The main
PCB was removed and setup on the bench using the Quasar matrix input adaptor plugged into
J4 to test:
- COIN 1 = 7 Credits
- COIN 2 = 1 Credit
Checking the Lazarian schematics confirmed that the cabinet only used the COIN 2 input and
the PCB was declared working pending further investigation in the cabinet. The spare PCB
swapped into the cabinet was coining up and playing OK, however.
prswan@gmail.com