Battle Bakraid PCB: Graphics Issue Advice

[djsheep]

Was offered a Battle Bakraid PCB from a trusted member of another forum, however when they went to fully soak test it, the graphics became garbled. It was apparently working fine up until this point. I've now been offered this PCB at a very cheap price.

From what I understand, there are 5 x EPROMS are for Sound and Gameplay and 7 x Soldered Graphics EPROMS. Sound and Controls apparently work fine. Obviously if I bought the board, my n00bness would check the ROMS first and go from there. But just wondering if anyone looking at these pictures can identify what would possibly be the fault. Same hardware as Bakraid, Garegga, etc. I definitely don't want to go near if there's possible issues with the customs.

Kindly paging @Hatsune Mike, @Apocalypse, @caius, @Hammy

 

[Hatsune Mike]

This board is probably trashed and won't work without some resources none of us have. The Xilnx CPLD near the front, similar to Batrider's TVRMCTRL, is the "Text VRAM Controller". It is responsible for multiplexing the nearby SRAM to act as a Color LUT, Text character data (tiles), and Text Layout/Attribute RAM. Often, it is a point of failure on these boards, generating garbage addresses or distorting the tiles it fetches. Finally, I believe this chip also fulfills the function of being a simple DMA controller to transfer to these regions.

There is no dedicated ROM for text; it is DMA'd from the CPU space IIRC into this text RAM. So, this is not a ROM or RAM issue. It is telling that the game is booting up and passing all its self-tests.

The big GP9001 custom won't be the problem here, but the Xilinx CPLD is locked and is as good as a custom. I made some weak efforts to get it dumped a while back, but that was shelved when it was clear that it's locked and attaching to the JTAG chain and just dumping it vanilla isn't going to happen. The devices do show up on the chain, however.

Since I expect the configuration itself is intact on the CPLD, we might be able to sent out a few samples to be RE'd in China to fashion a replacement.

Basically, if this board is < $200 and you want to consider using it for science, then that's all I can suggest.

I highly doubt this is just a reflow issue, but I'd be happy to be proven wrong.

It seems like someone always says it, so I'll just preemptively write that someone writing a mister core is not likely to help us repair boards with this IC having failed.

 

[djsheep]

Wow, thanks for your analysis. I've asked the seller if he wants to donate it to help find a replacement. Will contact you via PM if that's the case.

Cheers for helping me dodge what would have most likely been a bullet.

 

[Hatsune Mike]

You can also try setting the screen flip DIP and see if it magically looks different (or better?!) one friend had a board where it was perfect with the screen flipped, but garbled normally.

 

[Apocalypse]

Is the board populated with fujitsu TTL chips?
More than often they are fine the first 5mn after a long period of storage then start to crap in a domino effect.