24 Apr 2019

Wonder Boy - Sega 1986 (repair log)

Game was a conversion but not sure if official or not:
- game has 2 Sega daughterboards
- labels are handwritten
- program ROMs match the encrypted wboy romset

Anyway, sprites had horizontal bars over them:



I dumped the sprite ROMs but they checked ok, then reseated the associated daughterboard with no result.
So I followed the sprite generation circuit from IC to IC finding nothing obvioulsy wrong.
This is where I reached the 6 DRAMs used for sprites. They were AM2148 devices, known to generate a lot of heat, however two of them remained cold (IC13/IC33): to me that indicated those two chips were doing nothing.
I pulled them and installed fresh ones (much faster though being 35ns were original ones were 70ns) and sprite issue was cleared:




Lisenced... Clearly asking for a legal loophole.

I ended the repair by testing all inputs and sound and confirmed there wasn't any other problem with the board.
Game fixed.


17 Apr 2019

Ikari III: The Rescue - SNK 1989 (repair log)

Game got stuck on the memory test screen with palette error C4 and sprite error P15.
I pulled both RAMs (6116 type for the palette, 6264 type for sprites) but they both tested good out of circuit.
As you can't rely 100% on out of circuit tests I installed 2 fresh chips on sockets: errors were still present.
By probing the RAM chips them I could see some signals were floating: board had few points of corrosion and a couple of traces got severed.
I patched them all and then board booted into game but only the very top of the screen was visible:



I continued my inspection of the board, testing continuity of each dodgy point when I found pin 12 of the LS367 @D7 was floating.
For once schematics were available (quite unexpected since Ikari III is a bit of an uncommon game) and from them I learned missing signal was AB11 (address signal buffered through a LS244). Again I patched the severed trace and fault vanished:




Game fixed.

P.S.: This game seems to accept rotary controls through additional connectors on the board but is also totally playable with a joystick from JAMMA inputs.

10 Apr 2019

Knuckle Bash - Toaplan 1993 (repair log)

An other filthy board. But after a good bath in the garage sink it was much better:


Game booted with corrupted graphics, sound and controls were fine.





Since only part of the graphics were bad I thought of a ROM issue.
By having a quick look at the board layout I was sure graphic data were stored in the 4 big ROMs 3/4/5/6. I started by pulling ROM 5 to see if there was any change. There wasn't. Could I have found the issue so quickly?
Well, yes! Dump produced a file full of 0xff. I burnt a replacement in a 27C160 EPROM and graphics turned back to normal:




Game fixed.

3 Apr 2019

TMNT Turtles In Time - Konami 1991 (repair log)

I received this board as a gift, being warned it had a graphical fault.
Still, I was extremely grateful.

Game worked fine with the exception of jailbars over sprites:




One mask ROM had been replaced (flux used was probably not the best since it left a lot of residue, note to self: clean that) and few traces had been patched:





ROM test reported 2 bad:


Fortunately problem was simple, one via under one of the mask ROMs was corroded and connection between one data line from the mask ROMs and the sprite generator (custom chip) was lost. I redid the patching work and fixed the via by drilling a tiny hole in its middle and soldering the core of a very fine kynar wire in it, restoring connection between the two sides of the PCB:



Cowabunga! Game fixed.