Nov 072018

Received for repair this original Wonder Boy III – Monster Lair PCB (on Sega System16B hardware)

Board booted into game but sound was missing and some objects glitched :

I decided to troubleshoot for first the lack of sound.The amplifier was previously replaced as well as some electrolytic capacitors:

Using my audio probe I could figure out that the lack of sound was not caused by a fault in the analog circuit but it had a digital nature so I went to check the relevant circuit made of a Z80B CPU, a 6264 SRAM, a ROM and a YM2151 (plus a UPD7759 for speech samples) and I found that pin 13 (data line D5) of the RAM was silent as well as on the respective data bit of the Z80 and the YM2151 :

But it was toggling on the ROM (a 27C512 EPROM) which lies on ROM board :

 There was clearly a broken trace somewhere so I run a jumper wire from the RAM to the ROM:


This restored the sound.

Now the sprites issue.Relevant data are stored in eight 27C512 on ROM board :

I dumped the devices and my programmer complained about a bad contact on pin 9 of the EPROM @B2 and pin 16 of the one @B7:

At a closer inspection it turned out the pins were cut so they didn’t make a good contact into sockets:

I rebuilt the pins and this fixed the sprite issue.Board fixed.

 Posted by at 7:56 pm
Nov 042018

Received from USA this mint Contra PCB (by Konami), actually with japanese stickers and ROM set :

The board kept resetting in an endless loop sign that CPU was executing no valid code hence the watchdog circuit was triggered:

The two program ROMs were dumped as good so problem was elsewhere.Board was almost fully populated with Fuitsu TTLs which are well known to be prone to failure.When probing them with a logic probe I found a floating output (pin 11) of  a 74LS08 @17D:

This was confirmed also by logic analyzing and scope:

As schematics shows, the missing signal from pin 11 is called VRC and it should  be a control line routed to the custom ‘007452’ :

I removed the TLL and tested it out-of-circuit where it failed miserably :

I installed a socket and fitted a good IC:

The board booted into game with no further issue.Repair accomplished.

 Posted by at 6:05 pm
Nov 042018

Received for repair this original The New Zealand Story PCB (one layer hardware revision)

Board played fine except for this issue concerning background graphics, they had jail bars though :

All GFX data are stored in eight 1Mbit 28 pin MASK ROMs :

I dumped them starting from the one on the right labeled ‘B53-01’ :

And my programmer complained about a bad contact of pins 18-19 (which are data lines)

As said the device is a 1Mbit 28 pin MASK ROM which no direct replacement exist of, the only option is to adapt a 1Mbit 32 pin EPROM (better with non-JEDEC pinout), that’s what a made :

Perhaps not the best looking solution but it fixed board completely.Job done.



This repair inspired me so I made my own design to adapt a 1Mbit 32 pin TSOP Flash ROM to the 28 pin 1Mbit MASK ROM pinout:


 Posted by at 2:56 pm
Nov 012018

Received this original Mortal Kombat 3 PcB (manufactured by Midway) for repair: The Power-On Self-Test reported three bad RAM devices @U14-U36-U37 :

Board then went into game with bad graphics and colors:

The three RAM given as bad are two Fujitsu MB84256 (32K 8-bit) palette SRAM @U36-U37 and one Vitelic  V52C8128 @U14 :

As for the two palette RAMs I had no problem in finding spares and replaced them with two equivalent with same access time of 70ns:

The  one @U14 was tricky since it’s a multiport VIDEO RAM with 128K x 8 DRAM and 256K x 8 SAM (in SOJ package) rarely used on arcade hardware :


I could find three possible equivalents:

  • Toshiba TC528128B
  • Micron MT42C8128D
  • Hitachi HM538123

I was about the order the part online when I remember I had a dead NBA Jam PCB (which runs on similar hardware) somewhere.I digged it out and found it luckily used these RAMs :

I removed the bad chip:

And installed the spare:

This fixed board completely.Job done.


 Posted by at 11:27 am
Oct 292018

Received this Vapor Trail PCB for a repair :

Board had two noticeable issues, the sound samples were corrupted and sprites glitched like if part of them were floating on screen

I decided to troubleshoot for first the sound issue.A quick visual inspeaction revealed a 470uF electrolytic capacitor was missing in audio section:

But it was just the one which filters the +12V so reinstalling it didn’t cure the issue.

PCM samples are played by two OKI MSM6295 on main PCB while data are stored in two 27C010 located on ROM board:

I dumped the devices and they matched the MAME ROM files:

ROM_REGION( 0x40000, “oki1”, 0 ) /* ADPCM samples */ ROM_LOAD( “fj06”, 0x00000, 0x20000, CRC(6e98a235) SHA1(374564b4e494d03cd1330c06e321b9452c22a075) ) ROM_REGION( 0x40000, “oki2”, 0 ) /* ADPCM samples */ ROM_LOAD( “fj05”, 0x00000, 0x20000, CRC(39cda2b5) SHA1(f5c5a305025d451ab48f84cd63e36a3bbdefda96) )

But then I remembered I repaired another same PCB which had only one EPROM for audio samples whereas the other socket was empty.Other pictures found on the net confirmed this too like this one taken from ‘system11’ blog:

So I removed the one @1F and sound was fully restored.

Now the sprite issue.Hardware used the custom ASIC (in QFP160 package) marked ‘MXC 06’ which generate objects  :

We can see its pinout in the Oscar schematics:

Since the ROM board covers entirely the ‘MXC 06’ custom ASIC denying me any access to probe its pins I decided to replace it given also the fact there is almost nothing else involved in sprite generation  :

I took the spare from a dead Sly Spy/Secret Agent PCB and soldered it pin by pin with the iron tip:

When I powered the board up again, the glitches on sprites were gone so I could declare the board as 100% working.End of job.

 Posted by at 7:50 pm