A friend asked me if I could take a look at his dead Super Contra PCB which was a 2 boards version (the same game exist in a single board version where all the small mask ROMs on the sub board are replaced by bigger mask ROMs on the main board).
All I got was a black screen with no sound. Checking with a scope around the CPU area revealed there was almost no activity. The CPU is a Konami custom 052001, hard to troubleshoot as I couldn’t find its pinout online.
While checking for possible loose connections on some chips. I found that bending a bit this small chip near the JAMMA connector (an electromagnetic interference filter) was booting the game:
It was a bit loose. I desoldered it and found one of its legs broken. This was hidden behind one of these two small black ferrites. I repaired the leg and resoldered it.
Well, the game was now booting but graphics were a bit messy:
ROMs containing graphics and voices are located on the sub board.
I started checking this part as it has a pretty simple layout (basically mask ROMs with buffers) and noticed 3 mask ROMs with corrupted signals on their data lines. These 3 ROMs were all connected on the same bus to pins 2 to 9 (A bus) of a 74LS245 buffer @ location D1:
There are several other 74LS245 chips connected to other mask ROMs on the board and they all seemed to have “normal” activity signals. The B bus of the suspected chip (pins 11 to 18) seemed inactive and was connected to a custom chip on the main board.
I first tried piggybacking that 74LS245 @ D1 but nothing happened. The lines were still looking the same with garbage signal on bus A side, which could be caused by the possibly faulty chip itself. I desoldered it and it was tested bad on my IC programmer. I put a new one in place and the problem was resolved:
Here is what every signal on the A bus (pins 2 to 9) of the 74LS245 chip looked like with the faulty chip (left) then with a new chip in place (right). You can clearly see how bad the signal looked like on the left: