Apr 292019

It’s big, black and squared.We are talking of the ’85H001′, a custom IC you can find on some arcade PCBs manufactured by Capcom like Ghosts’n Goblins, Legendary Wings, Section Z, Trojan, Gun.Smoke.It looks like a module with audio functions, two versions have been made.A ceramic unmarked one :

And a plastic/epoxydic one marked ’85H001 5H’ or ’85H001 5I’ :

Technically speaking the ’85H001′ can be considered a nearly complete digital sound system, all typical components (CPU, RAM, address decoder, etc.) are embedded into it with the exception of the ROM and synth ICs which are external.Actually the ’85H001′ has been already reproduced by ‘Apocalypse’ who made an excellent thru-hole version :

You can read more about in his blog :


Having found recently  a bootleg of Ghost’n Goblins with a module replacement I decided to make my own reproduction and I made two versions.An ‘hybrid’ one with thru-hole Z80 CPU/RAM and surface mounted ICs  :

And a full SMT version with Z84C0006 CPU (PQPF44 package)

Here’s final testing of both on my original Gun.Smoke PCB:


 Posted by at 6:41 pm

Konami ‘501’ reproduction

 Reproductions  Comments Off on Konami ‘501’ reproduction
Apr 282019

The Konami  ‘501’ is one of the many custom ICs you probaby encountered when dealing with PCBs from this manufacturer.Most of times the chip (28 pin 600 mil DIP package) comes with scratched-off part name, sometimes not like on my Rush’n Attack PCB:

It can be found on not so many PCBs like shown on this useful spreadsheet (credits to ‘mattosborn’ on KLOV forums)

Actually there is already a reproduction of it which uses the ColinD CPLD 28 pin board (based on Altera EPM7064S)  of which ‘Porchy’ wrote the code (available here for download).I simply made what I call a ‘poor man’s’ version using simple surface mounted TTLs gates:

Here’s final testing on the Rush’n Attack PCB:

 Posted by at 5:06 pm
Apr 262019

Code name : ’86S100′.Nothing really ‘TOP SECRET ‘, just a little custom IC you can find on many Capcom arcade PCBs from pre-CPS1 era.To name few :

  • 1943
  • Bionic Commando
  • Black Tiger/Black Dragon
  • Mad Gear
  • Rush & Crash/The Speed Rumbler
  • Side Arms
  • Street Fighter

The chip is a 28 pin one with plastic package and pitch of 1.778mm :

Sometimes you can find it under different part name but same functions, like on a Mad Gear PCB :

This custom IC is involved in graphics generation being directly connected to data bus of the GFX ROMs as shown on Bionic Commando schematics:

Being a not really reliable part (sometimes you can find it literally ‘fried’) and quite hard to obtain as spare I decided to give a try to reproduce it.As usual I looked at how bootlegers re-engineered it and how the custom functions were implemented in earlier Capcom hardware.I succeeded in my purpose it but I actually had to make two versions of it because the original ’86S100′ can work in two modes depending on the logical state of its pin 1 (if hard-wired to GND or VCC).

Here’s final testing of both reproductions on a Bionic Commando and 1943 PCBs:


 Posted by at 7:29 pm

Teenage Mutant Ninja Turtles repair log

 PCB Repair Logs  Comments Off on Teenage Mutant Ninja Turtles repair log
Apr 222019

Got for repair from the States this Teenage Mutant Ninja Turtles PCB (actually the US 4 Player ROM set)

Board booted up with noticeable graphical issues, both sprites and backgrounds were affected as they were missing parts.Also colors were wrong, screen was blue tinted (hard to distinguish in the video and pictures though)

The palette circuit is made of two 2k x 8-bit static RAMs, a couple of latches (74LS273), some open collector buffers (74LS07) and finally three ‘052535’ RGB DACs that outputs each color to respective JAMMA edge connector pins :

A blue dominant means the RED color has some troubles so I went to probe around this part of circuit and found the 74LS07 @D23 (from Fujitsu, obviously…) with an output (pin 4) shorted to GROUND :

Chip failed in that pin when tested out-of-circuit:

Replacing the TTL IC restored the correct colors so I moved on the grahical issues.I launched a MASK ROM check which reported two bad devices @K4 and K6:

They are two of the four 4Mbit MASK ROMs that store sprites data:

I removed the first device @K6 :

I dumped it, the resulting buffer of my EPROM programmer was empty so device was really bad :

I launched again a MASK ROM check, the device @K4 was reported as good this time so the bad one @K6 was affecting it (data/address busses are shared)

I replaced the bad MASK ROM with a programmed 4Mbit EPROM (I used a Macronix MX27C4100)

Sprites were restored and  check no more complained :

But backgrounds were still missing parts :

This part of graphics is entirely handled by the ‘052109’ and ‘051962’ custom ASICs:

On a visual inspection I found a lifted pin on the latter:

I reflowed the pin and this fixed board completely.Repair accomplished.

 Posted by at 8:43 pm

Hammerin’ Harry (Irem M81) repair log

 PCB Repair Logs  Comments Off on Hammerin’ Harry (Irem M81) repair log
Apr 182019

Received for repair an Hammerin’ Harry PCB on Irem M81 hardware (it can be found also on M84).

Set is made of a top CPU board:

And a bottom VIDEO board:

The PCB simply booted to a static garbage screen:

The NEC V30 (uPD70116) main CPU was active on its data/address busses but when I probed the program ROMs I found a couple of stuck address lines :

Looking at hardware I figured out there are some 74LS373 used as latches to address the program ROMs :

Probing the one @IC39 revealed some outputs were floating while inputs toggling:

Once removed the chip failed the out-of-circuit testing :

I replaced the chip and the board booted up but on self-test showed most of times  a ‘RAM NG 8’ message:

Then it entered in game which was playable with sound but background colors were wrong :

The ‘KNA91H014’ surface mounted IC on bottom board @IC1 generates the palette for this part of graphics:

Its pinout/implementation from R-Type schematics :

I made a visual inspection and found a couple of lifted pins on a row:

I reflowed the whole IC with result that self-test no longer complained and graphics were fully restored:

The last issue I had is that some inputs were not wotking but I quickly pinpointed it to a missing custom resistor array near the JAMMA edge connector:


Having no spare I replaced it with a reproduction of mine:

Repair accomplished.

 Posted by at 3:02 pm