Shinobi Repair Logs
Repairer: channelmaniac
Symptom: Garbage on screen.
Found and repaired 2 cut/gouged traces and repaired bad hack solder job on the Yamaha DAC and LM324 audio op-amp.
Board was intermittently working and would stop/start working as the board was tapped or flexed. Narrowed it down to the CPU crystal oscillator. If you put pressure down on it the game would boot. Replaced the bad crystal oscillator and installed 2 missing NEC2501 ICs to finish the repair.
Symptom: Dead
Replaced 6 broken RFI filters on player 1 inputs. A couple were easy to spot as they were crushed. The others were found with a mulitmeter when running a continuity test.
Burned new unencrypted audio ROM chip from the MAME ROM set and replaced the encrypted sound CPU with a standard Z80B CPU to finish the repair.
Symptom: Dead
Game was totally dead. Game had NO custom CPU on it and no activity on the ROMs and the CPU HALT* like was held low, indicating a bus problem.
Checked the chips with the WCF (well calibrated finger) and found every ROM was getting hot and the CPU even hotter. Checked the ROMs in the EPROM programmer and found that all were bad. Burned new EPROMs from the Shinobi3 MAME set for all but the sound ROMs and burned the sound ROMs from the Shinobi2 MAME set. (Shinobi3 had encrypted sound and Shinobi2 had encrypted program ROMs)
Replaced the CPU and the board would then boot but had no sound. Checked the reset line on the Z80B CPU. It checked good. Checked the clock line and found it was missing. Replaced the crystal oscillator and then had a clock signal but not at the CPU. Replaced the 74HC74 clock divider IC and had clock on the Z80B CPU but still no sound.
Replaced the sound RAM to get the sound working and to finish the game.
The only thing I could think of was a power spike or (more likely) someone hooked up power backwards.
Symptom: Garbled graphics
Board took a hit on the corner where the PCB Foot is screwed down through the board. There are several traces that run through that area.
Repaired 2 damaged traces on the bottom of the board and 1 on the top then tested. The traces on the top board in that area carry sound data, the bottom connects to the video subsystem.
There are 8 traces on the top of the board but only the outer most of these was damaged. The rest were OK. However, these are LONG traces that go all the way around the outside of the board and are prone to damage. Here they are from the outside of the board inward - just in case it's needed for later.
E20, Pin 14 to CN4, pin A12
E20, Pin 13 to CN4, pin B12
E20, Pin 18 to CN4, pin A11
E20, Pin 17 to CN4, pin B11
E20, Pin 7 to CN4, pin A10
E20, Pin 8 to CN4, pin B10
E20, Pin 3 to CN4, pin A9
E20, Pin 4 to CN4, pin B9
Symptom: Dead
Repaired a broken trace on pin 10 of the CPU and tested. Repaired a broken trace to fix a graphics corruption problem. Replaced the dead custom sound CPU with a Z80B and reprogrammed the EPROM at location A7 with suicide-free code to finish the repair. Tested.
Repairer: Paul Swan
BOARD NO: 0122
GAME: Shinobi (on a System 16 B-type)
MANUFACTURER: Sega
DATE: 1987
HARNESS: SEGA-S
SYMPTOM (06/09/96)
No sound.
SOLUTION
Digital noise from the speaker suggested the amplifier was running and this was confirmed with a pulse injector applied to the pins of the Yamaha audio chip. Using a scope to look at the address and data buses at the sound EPROMS (A7, A8 & A9) indicated the sound processor was running but not responding to the game. Using a pulse injector on the data bus caused brief noise bites but nothing much else. On the advice of a fellow RGVACer, I ditched the custom sound module and replaced it with a standard Z80B and a new EPROM set. This fixed the sound completely. If anyone else needs the "Z80" images, mail me.