As often happens on this hardware the sprites had jailbars thru:
Board uses 4Mbit MASK ROMs to store GFX (sprites and background) data that are quite prone to failure.I launched a MASK ROM check that reported a bad device @K8:
I removed it and my programmer warned me about pin 13 (data line D0).I tried the dump in MAME and I could reproduce the issue :
I replaced the MASK ROM with an equivalent 27C400 EPROM :
Received from Portugal a batch of faulty PCBs for repair, there was among them this Bomber Man World (on Irem M97 hardware)
The PCB was working fine except for the sound, you could only hear some scratchy noise and also video output was affected showing some kind of disturbs:
This is scenario I faced many times during repairs and the cause was always the same : the electrolytic capacitors.In particular Irem boards suffers from the “brown ELNA” syndrome and this was not an expection.I went to measure capacitors with my ESR meter starting from the 470uF 25V one @C203 which filters the +12V routed to to sound amplfier.My meter didn’t display any reading meaning the measured ESR value was out of range (from 0 to 99 Ohm on my meter)
Probing the +12V line with a scope confirmed there was some ripple (no longer filtered by this capacitor)
The out-of-circuit testing confirmed the high ESR and bad capacitance :
Fitting a good capacitor restored loud and crisp sound.End of job.
Received from Spain this Cadash PCB (it turned out to be an undumped Spanish version)
Board was stuck on a garbage screen :
A clicking sound suggested that the watchdog circuit was active, infact both /HALT and /RESET lines of 68000 main CPU were going LOW and HIGH in an endless loop.At first glance I thouhgt the Taito custom ‘TC00220IOC’ (responsible of generating the master reset for whole system) was faulty since this is a very prone to failure part :
But I was wrong since I replaced it with no changes.Probing the 68000 CPU I found that data line D0 (pin 5) was stuck low, using a multimeter revealed it was shorted to ground:
Looking at solderside I traced the pin back also to pin 25 of a Fujitsu MB8422 Dual-Port SRAM (part of linking mode circuit)
I cut the trace and the short was cleared (so the I/O pin of the MB8422 was internally shorted).Finally board booted into game:
But I noticed two problems : the buttons of the Player 1 were not responding and sound had rustles, you can hear it from this video :
As for first issue, using a logic probe revealed the relevant pins of JAMMA edge connector were stuck LOW :
Inputs of an arcade board must be HIGH when not activated, they are usually pulled in this state by resistors.Indeed, tracing back the pins of the three P1 buttons from JAMMA connector lead me to some custom resistor network:
The array @RRC1 was the one involved so I removed and replaced it, this fixed the controls issue.As for sound, using an audio probe I figured out that the bad audio was coming out from the TL074 OP-AMP from Texas Instruments (which are really prone to failure, have replaced tons of them)
Received from Austria this faulty Kuri Kinton PCB, an obscure fighting game manufactured by Taito in 1988 :
Board is very compact, hardware platform is called ” Taito L system” based on the the ‘TC0090LVC’ (a custom Z80), an all-in-one CPU/system controller which does everything (game logic, tile handling, sprite handling, I/O handling, etc).At power up I was greeted by this scenario:
The game played almost blind, you could coin up but graphics were all scrambled.Ruled out the two GFX 4Mbit MASK ROMs, very few was left to probe.There are four 62256 SRAMs (32K x 8-bit) whose data/address busses are tied to the ‘TC0090LVC’ :
Probing them revealed unhealthy signals on data lines of some of them:
I decided to remove them all :
Actually only the ones @IC9 and IC10 failed the out-of-circuit testing:
