BOARD NO:     0065
GAME:         Time Pilot
MANUFACTURER: Konami
DATE:         1982
HARNESS:      KONAMI-S

SYMPTOM (21/07/96)
-------
Background split and duplicated. Spurious video corruption.

SOLUTION
--------
Moving the ribbon cable affected the video and changing it fixed the spurious video corruption leaving only split and duplicated background. The duplicated background effected only the text and ship fire, with the clouds and other objects unaffected. Using a pulse injector on the various ROMS on the larger board identified ROMS 1,2 & 3 as program, 4 & 5 as "clouds" and 6 as text. A logic probe verified that all the address and data of ROM6 were running normally. Working back, the most of the address bus for ROM6 was fed from 14H (LS273). All the inputs and outputs of 14H were running normally. 14H was fed from 13F (LS245) and then this fanned out to 11F, 11H and 6C. All these devices were running normally. Continuing to work back was begining to suggest more buffers and more fanout and getting me nowhere. The video RAM seemed to consist of a block of four 2114's all sharing a common address bus. A pulse injector indicated that these four RAMS controlled mostly everything in some way. All the address and data for the RAMS were pulsing normally according to the logic probe. The address for the video RAM traced back to a pile of LS157 mux's at and around J4. These mux's are four input and one arm was driven by the CPU address bus, one arm from some counters and an 82S153 and one from elsewhere that also formed an input to one of many LS86 XOR's next to the mux's. Checking all the LS157's with a logic probe suggested nothing suspicious - everything that should be running was. Moving on to check the LS86 XOR's revealed K3 pin 8 always low despite K3 pin 10 low and K3 pin 9 pulsing. Using a pulse injector on K3 pin 8 miraculously fixed the game. Flexing/tapping the board had no effect. The freezer spray had no effect. The heat gun had no effect. There seemed to be no physical cause and the game has run fine ever since(?). If it fails again, I'll know which device to change...