PDA

View Full Version : Moonwalker



Womble
26th March 2008, 09:46 PM
Thought I should put this in with the repair logs as it was only briefly mentioned in another thread but not somewhere where it might be helpful.

I bought the board off a guy who had got it in a consignment of games that were supposedly working. When he got it he found a some damage to one of the corners and the game would crash. He pulled all the chips that were socketed and consigned the board to his scrap pile. I saw it while I was round his place and bought it as a fixer. He put all the chips back on the board, plugged in the rom daughter board and $20 changed hands.

The edge damage turned out to be a single track that was broken, it looks very much like someone has tried to lever the ROM daughter board off and actually cracked the mainboard corner off. Either that or its had some weight on it at that corner and the metal mount has done the same job.

The board crack goes right through the via for this track, thankfully the damage was easily fixed and in an area where there isn't much else to damage. This corner no longer has the strength to support a mounting leg on it again so the loss of the use of that hole is no drama.

http://www.neoncherry.com/retrorepairshack/images/moonwalkerupper.jpg

http://www.neoncherry.com/retrorepairshack/images/moonwalkerlower.jpg

The game then didn't crash spectacularly, but it did do crazy things. The sound was missing, only a regular ticking noise from the speakers. The game would coin up and hang, if left in attract mode it would all look OK until the MJ sprites started to fight the bad guys. They would then crouch and glide around the playing area. Also there was some wide striping (15mm bands) in some of the graphics on the title screen. RAM/ROM tests were clear but the VRT test failed claiming IC68 was faulty.

After a lot of faffing about, looking for more track damage I ended up googling for details of the sega custom IC68, which is part number 315-5313. Didn't find much useful info about it itself, but did find a page that listed a large number of Sega chips for these boards as they all start with "315". There are 3 GAL chips that are also 315-xxxx in a bank of sockets near IC68. One of which was described as a VRT decoder chip, yet it wasn't the closest to the VRT chip IC68. No great dramas as none are overly far away from it, what did catch my eye was another 315 chip in this same bank that was a Z80 decoder chip - its a bloody long way away from the Z80!! More specific googling turned up that most of the GAL chips were in the wrong sockets. Each GAL chip has the IC markings denoting its type, they were in the right sockets for their chip "type", this is silk screened under the IC sockets. The have another number printed on top of them too, the 315 number that denotes the software that is held in them. Right chips, but software in wrong places.

As I suspected the Z80 decoder chip should be in the socket that's right next to the Z80 CPU.

Could never find a definitive list for the 6 GALs on this board, but after a bit trial an error with the few I could not get info in, here is where they should go.

IC65 = 315-5430
IC62 = 315-5375
IC74 = 315-5374
IC72 = 315-5389
IC80 = 315-5391
IC97 = 315-5390

Powered it up and everything kinda worked fine, it did lock up once or twice. There was a rather sick looking 470uF 25V C near IC68, it was badly wrenched to one side and one of its legs had dragged some of the guts out. I cut it off, put a tested salvaged C of the same value in its place and the board has been rock solid ever since.

Weird arse game tho, what WAS that man thinking???

Apologies for length.

Womble
2nd September 2008, 11:38 AM
Board started to get flaky again, would lock up when a coin was added, actually it played the 1st half of the coin-up sample then the screen scrambled. Turned out the sound eprom was duff, half the contents read as FF, so clearly some of the internal addressing circuitry has died. Replaced that ROM and now the game would coinup and start normally.

Board also would not get past the transistion from level 2 to level 3, its the 1st time that walls rollback to reveal a level inside a building. I thought that perhaps this computationally intensive effect was causing volt sag, and it turns out that the game would work fine if the 5V rail was cranked to up 5.7V, dangerously high.

I took a punt and bought the decrypted roms from www.segaresurrection.com and reprogrammed the 2 system ROMs, swapped the FD1049 encryption CPU with a stock standard 10Mhz 68000 and the game is again rock solid, even at borderline low voltage which is a very good sign.

I just wonder what else is going to go wrong with this board, its had a long line of problems to get this far.

hamish_nz
2nd September 2008, 11:42 AM
simple crack :lol

great work I love ready Repair logs