The game was working good except the sprites were all missing some lines
The board is using some unreliable drams 41464 from Fujitsu similar to the ones used on Konami Haunted Castle or CPS1 A board.
I bought some units from Aliexpress , put some handmade socktes and changed both to be 100% sure
Turned out only one of the two was faulty, in anycase the graphics were 100% fixed
I got this board for my collection as untested from an arcade operator
Board was the typical Konami pcb manufactured in Italy ( I believe Electronic Devices had a deal with them to import the pcbs and populate certain chips in Italy).
You regognize these boards because they have sound z80 made in Italy , soldered eproms and cheap sound potentiometer
Later on, beginning of 90s they started to put also general “E.Devices” stickers on eproms in place of Konami ones, but at least the eproms were socketed
Back to this board, it booted and the graphic was OK but there was a continuous metallic sound all the time.
Ruled out DAC and OP AMPS because usually when these fails, you have at least either FM or samples working, I started to probe the sound circuit logic following the schematics.
I noticed that the commands from the main CPU were correctly sent to the sound section but as soon as the Z80 had an INTERRUPT request, the CPU didn’t aknowledge it and blocked itself
In particular:
Signal IORQ didn’t go low, thus making the INT signal from the 74LS74 always LOW.
I proceeded to desolder the Z80 and tested it on another game with a socketed CPU and it confirmed to be bad.
While soldering the same Z80 by NEC used on proper Konami assembled boards, I also changed the sound potentiometer with an original Konami taken from another scrap game.
Game was fixed without any other issues.
Got this faulty pcb for my collection
Game was resetting over and over right after the first stating screen without even the possibility to see the diagnostic screen “RAM ROM OK”
The code was running and the watchdog was triggered by some event which I eventually found in a faulty 2114 SRAM @E15 on the video board
Marked in red on the actual pcb:
After changing it, the boot process went on until it displayed RAM and ROM OK but it reset again just before screen with the white grid.
At this time I was complete blind , missing any programming skills I couldn’t check what the code missed to go on.
I had a suspect on the sound part and after checking the clocks on the 8039 MCU I noticed it was completely missing.
The oscillator was working correctly and from the schematics I could see that the clock for the 8039 and the AY is generated using some 4bit binary counters 74ls293, very difficult to find on other boards.
Probing pin 9 of 74ls293@A12 it was stuck high while on pin 10 I had the clock.
I proceeded to test it out of circuit but was tested good.
Not sure about the reliability of the tester, I found a Starforce boot which use one 74ls293 and I installed it in a socket on megazone.
Still reset and no clock from pin 9
The output of the IC@A12 goes to pin 8 of a 74ls240, therefore I decided to test pin 8 of IC @D14 against +5V and it gave me a short.
After changing it the game finally booted without any other faults
Got this Devil World PCB as non working. This game is running on a Konami Twin 16 hardware. It is a two layer board with the first one devoted to CPU/sound and the second one to GFX/video. The two faulty/replaced chips were on the video board and are highlighted in red.
Luckily the board was booting and got stuck on the RAM/ROM test screen with ROM #18 reported as bad (chip #1 on the board picture above).
That ROM is part of a 4 ROMs row which contains the GFXs. They are 40 pins Konami chips and looking in MAME I could see that they were 512kb MASK ROMs. I tried dumping it with my programmer but was unable as it found one dead pin. Compatible EPROMs for these are 27C400 so I burned one and replaced it.
The game was now booting but with a color problem on some tiles.
After a deep look at the video board I could find a Fujitsu 74LS153 located at 10H (chip #2 on the board picture above) with floating outputs (while inputs had healthy pulsing signals). I piggybacked it and got everything corrected. I desoldered the faulty one, put a socket and replaced it. Board is now working perfectly.
Got another Konami game with the usual jail bars defect that is so common on these games
I ran again the maskrom test and it reported bad maskrom 8B.
After desoldering it and testing as 27c160, the programmer complained about missing contacts on several pins.
After replacing the markrom with a 27c160, the game was 100% fixed
