Hi everyone. I'm hoping I can get some help diagnosing a problem with a Mercs board (US version). So far all my work has been on the B board, because I currently do not have another Capcom game with a compatible A or C board.
The issue:
The story:
The game worked flawlessly at the seller's house. We transported the cabinet with the mercs board still mounted inside on a flatbed truck for a 2 hour ride. Upon booting up the game at home, the lines/jailbars appeared.
My guess:
The vibrations from the truck ride loosened a connection on the board, or perhaps a loose wire inside the cabinet brushed up against the board causing static electricity.
What I've tried, and what I've noticed:
- Disconnecting and reconnecting A, B, and C boards and doing a flex and press test on these boards. It seems like pressing or moving the C board changes these lines--sometimes better or worse.
- Cleaning and lightly sanding all roms on the B board. As I slotted roms back in one-by-one, booting up each time, the game had no lines until a rom would randomly trigger the lines. Then, the lines would not disappear until ALL graphics roms were removed. It's like the lines were stored in memory somehow. The lines do not appear to be isolated to one rom. Furthermore--without touching the board between bootups--sometimes the sprite lines are there, and sometimes they are gone. The background lines always remain. My supergun/power supply setup always runs at 4.99v or 5.00v, so I am confident it is not a power issue. Plus, it acted the same in the cabinet with a different power supply.
- Using a logic probe to compare signals on removable roms that have similar function or same chip model. The signals seemed to match each other when appropriate. I do not believe there are any broken pins or slots.
- Using a GQ-4x4, I dumped the contents of roms 30, 31, 35, and 36 onto my PC. I compared the data of the original roms to the mame roms, and none of the data is the same. Well, in some cases large chunks of it seems the same, but the way the data starts out is totally different. Is it because the software for the GQ-4x4 does not have the exact models of the chips? How can I reliably check if data is corrupted on the original roms?
All 4 of my original rom dumps did not work in mame. They either produced missing graphics with no game logic running, or the game froze completely, or just a blank black screen on startup. For reference, here are the chips I dumped and the closest models I could find in the GQ-4x4 software.
30:
model on chip: M5M27C100K-2 (Mitsubishi brand)
model in software: M5M27C100K
31:
model on chip: M5M27C100K-2 (Mitsubishi brand)
model in software: M5M27C100K
35:
model on chip: HN27C301G-20 (Hitachi brand)
model in software: HN27C301
36:
model on chip: TC571001D-15 (Toshiba brand)
model in software: TC571000D
If there is corrupt data, do I need to get new roms that match the exact same model? Is there any copy protection or anti-piracy issues I need to work around?
If anyone has suggestions please let me know. This is my first time trying to repair a board with this serious of an issue. I'll keep an eye on this thread, and either way I'll update with any progress.
The issue:
The story:
The game worked flawlessly at the seller's house. We transported the cabinet with the mercs board still mounted inside on a flatbed truck for a 2 hour ride. Upon booting up the game at home, the lines/jailbars appeared.
My guess:
The vibrations from the truck ride loosened a connection on the board, or perhaps a loose wire inside the cabinet brushed up against the board causing static electricity.
What I've tried, and what I've noticed:
- Disconnecting and reconnecting A, B, and C boards and doing a flex and press test on these boards. It seems like pressing or moving the C board changes these lines--sometimes better or worse.
- Cleaning and lightly sanding all roms on the B board. As I slotted roms back in one-by-one, booting up each time, the game had no lines until a rom would randomly trigger the lines. Then, the lines would not disappear until ALL graphics roms were removed. It's like the lines were stored in memory somehow. The lines do not appear to be isolated to one rom. Furthermore--without touching the board between bootups--sometimes the sprite lines are there, and sometimes they are gone. The background lines always remain. My supergun/power supply setup always runs at 4.99v or 5.00v, so I am confident it is not a power issue. Plus, it acted the same in the cabinet with a different power supply.
- Using a logic probe to compare signals on removable roms that have similar function or same chip model. The signals seemed to match each other when appropriate. I do not believe there are any broken pins or slots.
- Using a GQ-4x4, I dumped the contents of roms 30, 31, 35, and 36 onto my PC. I compared the data of the original roms to the mame roms, and none of the data is the same. Well, in some cases large chunks of it seems the same, but the way the data starts out is totally different. Is it because the software for the GQ-4x4 does not have the exact models of the chips? How can I reliably check if data is corrupted on the original roms?
All 4 of my original rom dumps did not work in mame. They either produced missing graphics with no game logic running, or the game froze completely, or just a blank black screen on startup. For reference, here are the chips I dumped and the closest models I could find in the GQ-4x4 software.
30:
model on chip: M5M27C100K-2 (Mitsubishi brand)
model in software: M5M27C100K
31:
model on chip: M5M27C100K-2 (Mitsubishi brand)
model in software: M5M27C100K
35:
model on chip: HN27C301G-20 (Hitachi brand)
model in software: HN27C301
36:
model on chip: TC571001D-15 (Toshiba brand)
model in software: TC571000D
If there is corrupt data, do I need to get new roms that match the exact same model? Is there any copy protection or anti-piracy issues I need to work around?
If anyone has suggestions please let me know. This is my first time trying to repair a board with this serious of an issue. I'll keep an eye on this thread, and either way I'll update with any progress.
