LIghtning Fighters repair log

Received for repair this Lightning Fighters PCB (a vertical-scrolling shoot 'em up developed and released by Konami on arcade in 1990)

Board was labeled with sprites and sound issues, indeed I got wrong colors on some sprites 

I already experienced this kind of problem on this and similar Konami hardware, the part involved was most of times the '053251 custom ASIC which is a priority encoder and sprites/backgrounds mixer 

 I went to check it with a logic probe and promptly found a floating pin (30)

This is an input to the custom IC so I visually inspected the board and found some exposed traces on solder side (soldermask was removed)

I checked them all and indeed the one routed to pin 30 of the '053251' custom IC was broken.I patched it with some 30AWG wrapping wire but sprites were better but not yet pefect.So I kept probing the '053251' custom IC and found a stuck low input pin (38)

 

With the help of schematics I traced it back to pin 112 (labeled 'OB08') of the '053244' sprites generator (along with its companion '053245').

I lifted ths pin and it was still stuck low meaning the custom IC was internally faulty.I took a spare from a Premiere Soccer donor board (actually the later revision '053244A', fully compatible and  interchangeable with the '053244') 

This fixed the sprites completely so I focused on the sound issue.The music was present but many sound FXs were missing or incorrectly played

Sound samples are stored in a 4Mbit MASK ROM (at location C5) and played by the '053260' PCM/ADPCM custom IC

 

I performed some MASK ROM checks that most of times the device @C5 was reported as bad

Also the sound test was silent, I could not hear anything when I sent commands to play the samples

Sure enough I removed the 4Mbit MASK ROM 

I read the device in my programmer and surpringsly the dump matched the MAME ROM file but I decided anyway to replace it with a programmed 27C400 EPROM 

This was the right move since all sound FXs were totally restored.Board 100% working again and another successful repair.

Konami '0005294' reproduction

After the '0005291', '0005292' and '0005293' another custom IC from Konami GX400 arcade hardware has been reversed., the '0005294'.This part shares the same QIP64 package of the '0005291' : 

Technically speaking it's the sprite engine of the entire system, here's pinout from Nemesis schematics :

The approach for making a replacement was the same of the other GX400 custom ICs, having a ready footprint for QIP64 package (previously realized for the '0005291') made things easier :

After some code adjusting I got the reproduction perfectly working :

Another step to a full remake of Konami GX400 hardware, stay tuned.

'NMK-903' reproduction

Thanks to the brilliant job of 'Furrtek' and 'Recreativos Piscis' another custom IC has been reproduced and preserved.

The 'NMK-903' is a small custom IC (in QFP44 package) found on arcade NMK PCBs from the 90s: 

Tracing and schematics of internal structure was done by 'Recreativos Piscis' starting from die shots made by 'Furrtek' who decapped the chip, too :

I designed a proper board (powered by a small CPLD) as replacement of this custom IC using castellated half holes :

The reproduction worked fine at first test :

A big thank again to 'Furrtek' and 'Recreativos Piscis' for their unvaluable work.

Konami '0005291' reproduction

After the last reproduction of the Konami '0005293' I reversed another custom IC from GX400 hardware, the one marked '0005291'.This part comes in an unusual 64 pin quad in-line package (abbreviated as QIP-64 or QIL-64) :

It functions as tilemap generator, here's snippet from Nemesis schematics :

After studying the behaviour of the original custom chip and gathered all info about it found on the net I started my task for create a replacement.This was quite a pretty challenge because I had also to recreate the original QIP footprint.After some time I developed a board powered by a CPLD :

The reproduction worked flawlessly at first shot :

Another step to a full re-engineering of the Konami GX400 hardware, stay tuned for next project.

War of Aero - Project MEIOU repair log

Got for repair this (never seen before) War of Aero PCB, a shooter released by ALLUMER in 1992.

Board was in very good condition :

But it was dead, no sign of life on power up.As first thing I went to check with a logic probe the 68000 CPU (in SDIP64 package) and found the clock input was stuck low :

I traced the clock pin back to a 74HC368 that drives the 16MHz signal generated by the nearby DIP14 oscillator :

Probing the output of the oscillator with a scope confirmed it was dead, no square wave from it :

I promptly removed the part :

And installed a spare :

This brought board back to life :

No other issues found so another board succesfully repaired.

Konami '0005293' reproduction

Another custom IC from Konami GX400 hardware has been reversed.After the '0005292' I reproduced time ago I took into account the '0005293' that can be found always on VIDEO board of this arcade system.  

It shares same SDP64 package :

Technically this custom IC is a priority handler, pinout can be found on Nemesis schematics :

For my task I made use of Salamander bootleg (where this and the other custom ICs have been reversed with the use of 74 series logic gates ) and gathered all info found on the net about the '0005293'.Prototyping was encouraging so I came to this replacement :

After some adjusting of the code and adapting the timing to the hosting hardware the testing of the replacement was successful on my boards :

X-Men (Konami) repair log

Recently I've been sent an original Konami PCB for repair.The board was not in mint condition, it was heavily reworked, a lot of ICs (RAMs, ROMs, TTLs) were socketed and replaced.

Board was marked as working but with missing sprites, that was indeed what I got when I powered it up :

I entered in TEST MODE and performed a MASK ROM check which reported all the sprites devices as not good :

Object are handled by two custom ASICs (in QFP120 package) that work as pair, the '053246A'  (which generates the addresses to the MASK ROMs) and the '053247A' (which draws the sprites) :

Looking at schematics made by 'jotego' from MiSTer (thanks to him!) I checked the connections between these custom ICs and the rest of the board, everything was correct.But when I went to probe the '053247A' I found some stuck outputs :

These are the color and priority bits sent to the '053251' (the priority control custom IC) :

The '053247A' was most likely faulty so I decided to replace it, luckily I had a donor board to take the part from. 

Desodered the defective part with hot air :

Installed the spare :

The sprites came back :

Another successful repair.