IGS PGM Mainboard Repair Log

dbstallman

Member
Joined
Jan 9, 2006
Messages
738
Reaction score
12
Location
Rock Hill, Missouri
Last week I picked up a non-working PGM mainboard off Ebay. The board had color issues most of the time. In test mode, the colorbars would display correctly, but for almost all of the gameplay everything would be tinted yellow.

Here is a photo of the video from the auction :
badvideo1.png


The RAM test failed on RAM 5, so I knew (kind of) where to start..... I just couldn't find any documentation about where RAM 5 was on the PCB. Looking around in the MAME driver and through a thread started by XianXi at neo-geo.com (http://www.neo-geo.com/forums/showthread.php?220574-PGM-Tech-Question) I was able to determine that the video RAM was the three SOJ chips in this area :
imag1009.jpg


In the neo-geo.com thread it was speculated that U21 and U22 were the main video RAM with U30 being for the pallette (this picture is from after I replaced U30, originally it was also a Winbond W24257AJ-8N) so I started by replacing it with a spare SOJ RAM I harvested from a dead Namco System 12 PCB.

I booted the board into test mode while the chip was off to see if it was really RAM 5, but the picture wouldn't reliably display. From what I could make out, it is probably RAM 2. After replacing the RAM there was no difference in the display and test mode still failed on RAM 5. Since the board was silkscreened for a 256K RAM in that position, I decided to just leave it there even though it was larger than the old RAM.

Next I removed U21 and fired the board up with the RAM from U30 pressed into place (I know, not reliable) and it did pass the RAM test all the way through RAM 5! I figured I had it solved and soldered the chip in place.

Unfortunately when I powered it back up RAM 5 failed the test again. That left only U22, so I pulled it and did another test with the RAM off the board. This time I was able to clearly see all the graphics and the test mode passed RAM 5 again. Hopefull, I soldered the RAM from U21 in and fired it back up.. This time it was a complete success! The test mode passes all RAM tests and the gameplay is perfect!

Here is what I think we can learn from this : RAM 5 is both of the chips at U21 and U22. XianXi had mentioned in the other thread that the RAM was mapped together in a rather unusal way and I suspect this is the result of that. You guys can correct me if I'm wrong.

Anyway, I hope this helps someone else since there seems to be very little info on the net about this system.

Brian.

Edit : I don't know how I managed to put a "Thumbs Down" icon in the title... If a moderator could remove or change that to a smiley or something I would appreciate it.
 
Last edited:
Nice work Brian. Like Brian said, there isn't much info on the net about the PGM. I hope sometime in the near future, some of you tech gods here on KLOV can help plot everything out on the hardware.

My experience with the PGM has usually been RAM problems with the occasional 27 error (Clock signal).

It'd be cool if someone can help plot the RAM locations out and what not.
 
I got a problem with a Gladiator cart. When you power on the system it says it is waiting for the cart to initialize and shows an address. First time I powered it on it said 0xFF and now it shows 0x6D and doesn't do anything else. I thought it could be the oscillator so I swapped it out and no change at all.

Already verified the roms, so it's either the 2nd CPU, RAM or TTL chip. Of the TTLs I would think the only one that could contribute to this problem would be the LS04 since they are usually used to disperse the clock signal. I haven't tried to swap it yet.

These carts are hard to work on because the mobo is a side loader and it's impossible to get a probe in there.
 
Back
Top Bottom