Konami (Sunset Riders) PCB resetting after (succesful) ROM check

[frsj8112]

I would check the CPU RAM (or work RAM).
I had a Sega Wonder Boy 3 PCB on 16A hardware that would boot and I could coin up and push start, but as soon as the first stage started to appear on screen, the PCB resetted.
It was due to a faulty CPU RAM.

Funny thing is that in attract mode, other stages could appear on screen, but as soon as the first stage was about to appear, it would also reset.

 

[ShootTheCore]

You'd start soldering on the board before using a logic probe? What's the reason for this priority?

You can piggyback a new RAM chip on top of the old RAM chip without soldering. But I was thinking through things this morning and realized that the startup diagnostic runs from Work RAM so it’s probably OK.

So, before you mess with the RAM, I did a little research and found that there’s an easy way to disable the watchdog reset circuit on that game. If you close the JP jumper next to the JAMMA connector then the board will no longer auto-reset. Give it a try and see if the game boots up past the startup diagnostic.

 

[Sumez]

Give it a try and see if the game boots up past the startup diagnostic.

I should probably have mentioned that I already tried this (took a hint from the repair log I linked previously), and with the watchdog disabled the CPU stops resetting, but it will be stuck on the test grid pattern that appears slightly before the game "boots up".

Here's a video someone uploaded of the PCB version I have, you can see the grid right before the copyright message at 0:07:

 

[ShootTheCore]

OK, so if it crashes on the startup grid with the watchdog disabled, I more strongly suspect the two Work RAM chips. Here's why:

1) If there was problem with the address lines to the Work RAM, then it probably wouldn't even get that far. It would have crashed while loading up the the startup diagnostics. That said, there could be an issue with one of the high-order address lines for accessing the end of the Work RAM bank, so don't rule out addressing entirely.

2) The grid is displayed while the diagnostic is unloaded and the actual game is loading up. The bootup diagnostic only uses a small chunk of Work RAM while the game fills up the entire RAM. Therefore I suspect that there's a bad block of Work RAM outside of the space that the diagnostic uses.

3) The diagnostic does check the Work RAM integrity, but I doubt it does a thorough check since it does it so quickly.

All of this is speculation of course - it's difficult to troubleshoot electronics remotely. But from the symptoms, I'd try socketing and replacing those two Work RAM chips.

 

[XianXi]

I’ve had a PCB before where it passed the rom check but when I pulled the rom to verify it, it failed. Replaced it and it worked perfectly after that. I told this to RJ before and I told him that it doesn’t make sense but it’s true.

This is the kind of stuff I wanted to hear. Was it a Konami board also?

Sorry missed the reply, yes I had it happen on both a TMNT and Sunset but my problems were graphical.

If it’s resetting, check the ram if there are any Toshiba 2018s, those always give me problems.

 

[Mitsurugi-w]

I had this issue with an xmen I just repaired. The problem was a bad through hole on the sound module. I had to trace every pin to make sure it had continuity down the line until I found one pin not connected to its trace. Patched it and it booted up.

 

[XtraSmiley]

I had this issue with an xmen I just repaired. The problem was a bad through hole on the sound module. I had to trace every pin to make sure it had continuity down the line until I found one pin not connected to its trace. Patched it and it booted up.

That was also the issue with my X-Men as well, wasn't it? My friend Bryan ended up finding and fixing it.

 

[Mitsurugi-w]

Might have been. Corrosion from cap electrolyte. Top and bottom pads looked good but corroded inside the hole.