Mar 202019

Got from USA this original Contra (by Konami) PCB for repair :

Board simply booted to a black screen, time to investigate then.A closer inspection revealed the 3.58MHz oscillator was missing:

This provides clock for the YM2151 IC as you can see on schematics:

I fitted it but obviously board was still dead so I focused on the main CPU ( an HD63C09) dumping the two program ROMs:

Analyzing dumps revealed that somebody installed another sound ROM (like the one @12A) in the socket of the program ROM @18A, you can see in above picture that indeed there are two devices with same ‘633E01’ label.I programmed an empty 27256 EPROM but board was still not booting.Looking better at PCB I noticed the ‘007593’ palette custom IC @1F was damaged :

The package was really cracked isolating pin 24-28 from the rest of its internal circuit :

The custom ‘007593’ exchanges 8 bit of color data with main CPU and other devices through pin 21-28 as per schematics :

This did make sense so I promptly removed the custom and installed some 1.778mm headers:

As some of you may remember some time ago I made a reproduction of the ‘007593’ :

Konami ‘007593’ & ‘007327’ reproduction

So I assembled and installed a unit :

Board booted into game with no further issues.Repair accomplished.

 Posted by at 10:20 pm
Mar 182019

Received from Portugal a cocktail version PCB of Indiana Jones, a real “holy grail” given its rarity.The peculiarity is that it runs on a different hardware than usual Atari System 1.Board is a two stack one made of a ‘main’ PCB :

And a ‘cartridge’ PCB :

All I got when I powered it up was a flashing ‘NO CARTRIDGE’ message on screen, watchdog was active and board resetting in an endless loop:

I started my troubleshooting with some preliminary things like reseating chips and ribbon cables.This worked in some way because it shutted down the watchdog and put the board in a static red screen :

Despite this state I was able to enter into TEST mode and perform all checks that were successfully carried except for the sound test which failed reporting this error :

The sound CPU is a 6502 located on cartridge PCB:

I replaced it with a good one and board booted into game.All seemed to be working fine, game was full playable but in a comparison with MAME I found that speeches were missing on my board

Here is how it should play:

I went into TEST mode and launched again a sound test.Music chip (YM2151) and Effects chip (POKEY) were tested good whereas the Speech chip test was silent:

The Speech chip is a Texas Instruments TMS5520:

Most of pins were inactive so,having a spare I replaced it but with no luck.Looking at schematics the TMS5520 exchanges data with the SY6522A (VIA, Versatile Interface Adapter)

I could not see anything else involved in the fault so I ordered the part which came into mail after some time:

I installed it and speeches were restored, board 100% fixed.Repair accomplished.

 Posted by at 4:07 pm
Mar 112019

Some time ago I bought a lot of original faulty boards, there was among them a Taito PCB which at first gance I could not identify:

I could not power it up because there was a dead short between GROUND and +5V :

So I dumped some ROMs and it turned out to be Extermination, a vertical shoot’em up game (that runs on “The New Zealand Story” hardware) released by Taito in 1987.Someone previously tried to troubleshoot the board and cut one terminal of a zener diode thinking it was shorted but this was for +12V line protection whereas there is another zener diode to protect +5V which was actually bad causing the short to GROUND:

Once replaced the zener diode and cleared the short I powered the board up and it booted into game but sound was missing and colors were wrong with a dominant blue :

A blueish image means the RED color is missing or has some troubles.This was confirmed by the logic probe, the signal was indeed stuck low :

I could trace the JAMMA edge connector pin of RED back to a SIL component marked ‘RO-012’, a typical R/2R resistor ladder used as RGB DAC (one for each color).A closer inspection revealed the part was mounted backwards:


Reinstalling it with right orientation restored the RED color but image was now purplish :

This meant the GREEN color had some troubles.I located the relevant ‘RO-012’ DAC near the sound amplifier and found that some pins were shorted by a solder bridge:

Removing the bridge finally restored the correct colors :

Now the lack of audio.From signs on solder side I noticed someone previously reworked the sound section replacing the amplifier (an Hitachi HA1388) and potentiometer (and perhaps also capacitors)

Checking the +12V  confirmed this voltage was present on board but not on amplifier when I put the black probe of my multimeter on the two GROUND pins of the HA1388 (pin 4 and 9).At the end it turned out that who replaced the amplifier and potentiometer managed to lose their GROUND connection likely ripping off the rivets from the holes (power pins are always harder to desolder due to presence of internal planes).I restored the connections in this way :

Sound was back again.Board 100% fixed and job done.Before archiving this repair I took the chance to see if some parts (simple ones, not digital ASICs) were worth to be reproduced.I chose the aforementioned ‘RO-012’:

And the ‘X2-003’ :

The first, as said above, is an R/2R resistor ladder acting as RGB DAC, the latter is a capacitors/resistors array used for inputs but, unlike all the others, comes in a DIP16 ceramic package hence quite fragile.Reproducing them was straightforward :

 Posted by at 8:53 pm
Mar 112019

Again my Defender boardset had developed another fault.

I had already repaired here:

Defender repair log #1

and here:

Defender repair log #2

This time after the screen “Initial tests indicate all unit OK” it rebooted in an endless loop.

The diagnostic LED were flashing two times correctly before resetting, so I thought it was the ribbon cable.

After spraying and moving the cable , game still reset itself after the initial test but after a while it didn’t produce anymore the classic startup sound and all the LED were always stuck lit.

The 6821 PIA on the romboard should send an \IRQ signal to the main CPU after the ram test but the signal was missing, that’s why the watchdog kicked.

Also the 4 LED stuck lit is an evidence that the PIA died.

It’s known to be not very reliable, so I decided to desolder it , put a socket and test the only one PIA I had for spare which luckily made Defender to boot again.

Game was  fixed without any problems ( for the moment )

Mar 092019

Got for repair from USA this Pac-Mania PCB on Namco Sytem 1 hardware.System is made of a larger ROM board:

And a smaller CPU board:

It booted to an ‘EEPROM ERROR’ message screen followed by some digits:

The EEPROM is a 2816 (2k x 8-bit) device located on CPU board:

I didn’t try to replace or swap this EEPROM since I knew it was not the cause of error.From my experience this is due to another IC that goes bad very often.I’m speaking bout the custom ’64A1′ located on ROM board @M4.Mine, indeed, was marked with a pencil as ‘BAD CHIP’ :

Technically the custom ’64A1′ is an HD63701 MCU in disguise with the exception of two custom opcodes.I already explained in a past post of mine how to replace it with a standard HD63701 MCU :

Namco System 1 custom ’64A1′ replacement

It’s what I did and this was enough to fix the board completely.Job done.

 Posted by at 7:36 pm