Extermination repair log and ‘X2-003’ – ‘RO-012’ reproduction

 PCB Repair Logs, Reproductions  Comments Off on Extermination repair log and ‘X2-003’ – ‘RO-012’ reproduction
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
Dec 062018

The ‘UEC-01’ and ‘UEC-02’ are two custom ICs you can find on most Seibu/TAD Corporation PCBs also marked as ‘HB-1’ and ‘HB-2’.They are ceramic modules in SIL/SIP package.Here’s how they looks like:

The first handles I/O, the latter drives coin counters/lockouts.Reproducing them was straightforward given the simplicity of their internal circuit so for sake of completeness I took into account also the ‘UEC-02’ although nowadays few people use coin counters:

Testing on a Blood Bros PCB :


The ‘JK-03’ can be found, instead, on Jaleco MegaSystem1 hardware, it accomplishes the same function of the ‘UEC-01/HB-1’ handling the inputs.It comes in form of a SIL/SIP ceramic module as well (with different pin counting and pitch though)

Having no part to analyze I used some available schematics in order to reproduce it:

Due to the lack of a working PCB I could not test it hence volunteers are welcome!

 Posted by at 6:38 pm

Silkworm repair log and Mitsubishi ‘MA7053’ reproduction

 PCB Repair Logs, Reproductions  Comments Off on Silkworm repair log and Mitsubishi ‘MA7053’ reproduction
Sep 282018

I had this original Silkworm PCB (manufactured by Tecmo in 1988) in a trade some years ago :

I didn’t know the status of the board since I never looked at it.Quite confident I powered it up but I was immediately disappointed, a solid static blank screen was all I got:

While I was visually inspecting the board I noticed the 28 pin brown SIL hybrid module marked ‘MA7053’ was a bit wonky:

I slightly wiggled it and it fell off:

I installed some 1.778 female pin headers on the PCB:

and patiently soldered the component on a strip of correspondent male headers:

In this way I got the board booting but with severe graphic faults.Color were wrong, sprites flashing (it’s hard to capture this issue with a camera), vertical lines through screen :

I decided to troubleshoot the sprites issue for first.Studying the hardware I figured out that the line buffer is made of twenty 4164 (64K x 1-bit) dynamic RAMs located on VIDEO board  :

Some of them were extremely hot to touch and many showed stuck bit on output:

I pulled them all one by one:

Nine of them failed the out-of-circuit testing:

The graphics were correctly drawn now but the colors still wrong :

Tracing the three colors back from JAMMA connector I figured out the final part of the RGB DAC circuit where I noticed the lack of three resistors @R10-R8-R12 :

I compared my PCB with some pictures online and I had confirm that the three resistors were really missing on my board:

I didn’t know the correct value of these resistors and schematics were not available so I looked at Rygar ones which runs on similar hardware, they were 120 Ohm as part of the R-2R resistor ladder circuits used as RGB DAC:

I installed the resistors :

This fixed board completely:


  • Mitsubishi ‘MA7053’ reproduction

After repaired the PCB I thought this was a good chance to study a replacement of the ‘MA7053’ custom SIL.As often I do in my reproductions I looked at how possibly the custom was re-engineered.I could find two replacement daughterboards.One used on a bootleg PCB :

The other one was from an original Tecmo board (pictures kindly provided by ‘monsterlair’, thanks again to him)

Design were slightly different but they both have same functionality.Technically speaking the ‘M7053’ provides interface between the Z80 main CPU data bus and video memory latching data too.After figured out schematics of the two daughterboards I re-engineered them with surface mounted devices ending up with this result :

Installed on PCB for testing:

Both designs perfectly works on my newly repaired Silkworm :

The ‘MA7053’ is used for sure on these PCBs:

  • Gemini Wing
  • Rygar
  • Silkworm/ Back Fire

But it could be present also on other Tehkan/Tecmo boards so any addition is welcome from all of you arcade collectors/enthusiasts.

 Posted by at 11:15 am

Konami ‘503’ reproduction

 Reproductions  Comments Off on Konami ‘503’ reproduction
Sep 262018

The Konami ‘503’ is a custom chip we can find on some PCBs from this manufacturer.It comes in a DIP40 package with its part number scratched off in typical Konami style :

You can refer to this spreadsheet (courtesy of user ‘mattosborn’ on KLOV forums) for a list of boards (all with Konami Classic pinout) that carry this custom :


Technically speaking, the IC takes care of part of the sprites handling (along with the other custom ‘502’)  as schematics shows :

We already have a modern replacement of the Konami ‘503’ thanks to the wonderful work of people over FPGAArcade forum, this is the thread of reference :


What I made is a “poor man’s” reproduction of this custom chip with no use of CPLD or other complex programmable logics but I simply re-engineered with surface mounted devices the replacement daughterboard (silkscreened ‘KC001’) that Konami used in place of the IC :

Here’s the result:

It works fine on my Track & Field and Kicker PCBs and should do the same in all other boards that use this custom IC.

 Posted by at 10:08 pm

Vigilante repair log and Irem ‘RGLD8M472J221J’ – ‘RGSD10L471G’ reproduction

 PCB Repair Logs, Reproductions  Comments Off on Vigilante repair log and Irem ‘RGLD8M472J221J’ – ‘RGSD10L471G’ reproduction
Aug 152018

Got this Vigilante PCB (by Irem) in a trade some years ago:

Board was dead, stuck on a purple static screen:

As usually I started my troubleshooting with a visual inspection and noticed severe corrosion in some areas, especially around the 3.579545 MHz oscillator:

Analyzying the main Z80 CPU revealed no clock on pin 6.This signal was present on oscillator output but then was lost when routed to the inverter (a 74LS04 @IC30) :

Replacing the TTL restored clock on Z80 but board was still dead.Probing the /RESET pin 26 with an logic analyzer showed an unhealthy signal, there was no proper transition from LOW to HIGH but only some oscillations :

/RESET signal is generated by the MB3771 voltage monitor @IC31.I promptly replaced it:

In this way proper signal was restored:

But main Z80 was still inactive, data/address busses were silent as well as control lines.This lead me to think the CPU was faulty so I removed it:

Testing it in another board confirmed it was really bad.With a good CPU the board finally booted up but the sprites were mostly absent, I could see only some parts of them randomly flying over the screen:


After some time spent to check different things I pinpointed the fault in a bad interconnect ribbon cable.For safety I replaced both of them:

Board fully working again and a quite enjoyable game added to my collection.

The repair was accomplished but, as always I do, I visually inspected the board looking for some candidate parts for a reproduction and I spotted two possible ones.The first is marked ‘8M472J221J’  (‘RGLD8M472J221J’ on manual parts list)

It’s nothing more than a custom resistor network used for inputs, you can find it also on other Irem hardware like M72, M92 and M107.Here’s snippet from R-Type schematics:

I reproduced it this way  :

The latter, marked ’10L471G’ (‘RGSD10L471G’ on manual parts list)  is interesting since it’s a R2R  resistor ladder used to convert to analog the 5 palette digital bits outputted by the surface mounted ‘KNA91H014’ custom.You will find always three of them (one for each RGB color) coupled to one custom.

Its implementation on schematics:

From my experience and other references too the original part is not really reliable (it will crack or burn) so I reproduced it as well:

Testing both reproductions on the repaired Vigilante PCB:

 Posted by at 11:15 pm