Mar 182014

I bought an untested Nintendo Vs PCB a few months back and it turned out to be faulty. I decided that I had nothing to lose in attempting to fix it, and finally, after much effort, I’ve managed to get it working!

The PCB started out by displaying a blank screen:

Nintendo Fault 1

Or by displaying an out-of-sync type screen like this:

Nintendo Fault 2

First off it was important to confirm that the CPU and PPU (which is like a security chip) was functioning correctly on the board. A fellow arcade collector came across a document by the University of Illinois that gave me loads of information about the pinouts of both these chips, and this was a good starting point.

So I checked the CPU reset pin (3), clock pin (29), and the PPU clock pin (18) with a logic probe to confirm they were working correctly. I also checked the address and data lines of both, and everything looked fairly good.

I removed the 6116 FP Series RAM from positions 1E and 6E on the PCB and replaced them with sockets with 6116P-2 RAM. Another collector came up with an idea that the RAM in these positions could be replaced with NVRAM. That way we could power up the board, let it run for a few seconds, power it down, remove the NVRAM module and read out the contents of those RAMs in an eprom reader. It would then be possible to compare the contents to the MAME equivalent of that game in the MAME debugger, and that would let us know whether the CPU was doing what it should with regards writing the correct data to those particular RAMS. It would also confirm if the CPU was executing code correctly and ultimately would eliminate it from the problem.

Before giving the NVRAM idea a go I checked the RAM at position 6E and found out that pin 21 was dead. This was the /WE (Write Enable) pin and should be pulsing on the logic probe to indicate the CPU is reading from, and writing to the RAM. Following it back to its source on the schematics it came from pin 8 of the NAND gate at position 5B. Sure enough the NAND gate at 5B looked broken. Pin 8 was dead. I piggy-backed it with another LS32, and pin 8 started to pulse and look correct, and so did all the pins now on the RAM.

I wanted to find out what was causing the out-of-sync image on screen I was seeing. I located an area on the schematics which centered around an LS04 at position 1L. Pin 5 was not acting as expected on this chip. I piggy-backed it with another LS04 on top of the broken one, and it made no difference. I even removed the transistor at position Q8 to determine if that was causing the fault and that looked ok, so swapping the LS04 out with a new one rectified the problem, and pin 5 looked healthy again.

The next part was a bit of luck really. I could see that there were issues with the RAM at position 8L. Looking at the schematics I traced this back to the LS157’s at positions 5K and 6K. I piggy-backed both and the game sprung into life! I removed the piggy-backed chip at 5K and the game continued to run. If I removed the chip at position 6K, it killed the game, so I knew this was at fault.

So I now have a functioning game!

Nintendo Working

Obviously from the picture above it looks like the game still has problems, but this is potentially related to a partially faulty PPU that I’ve used for testing. I will test the PCB with a fully working game to ensure everything is now working correctly.