Feb 062019

Some days ago I wanted to test some RAM chips from an arcade PCB using my B&K Precision 560A tester :

B&K precision 560A IC tester

For the uninitiated this piece of equipment allow you to do out-of-circuit and in-circuit testing of many ICs like TTLs,RAMs and ROMs thanks to an extensive internal IC library.I often use it during my repairs but in that day it suddenly died, no sign of life, nothing came up on display when I powered it up.For first I checked the main fuse and it was good so I decided to open the unit:

Its internals

Having no schematics I started my troubleshooting from upstream checking the big transformer just after the AC input, voltages were present on its secondary coil.Following the path I figured out the AC voltages are rectified by a circuit on one of the PCBs.In particular the rectified +12V is then regulated by a couple of ‘LAS1605’ fixed +5V voltage regulators (in TO-3 package) mounted on a big heatsink for providing +5V to the circuitries on the other boards:

Two LAS1605 fixed +5V voltage regulators in TO-3 package

I probed the inputs of both and +12V was present:

As well as +5V on the output of one regulator:

But the other one was giving only +1.706V on its output, too few for correct functionality of the logics on PCB:

I looked online for buying a ‘LAS1605’ +5V voltage regulator or a compatible one but found only few at not really cheap price.Then, as a last resort, I asked my local electronics shop, they surpringsly had in stock some TO-3 ‘MC7805CK’ which is a perfect drop-in replacement (although it can deliver up to 1.5A output current against the 2A of the ‘LAS1605’ but this is not really an issue since I don’t think the tester can draw more than 1.5A during normal operation)

MC7805CK used as replacement

I pulled out the suspicious part:

The culprit in all its loneliness…

I swapped the spare in and was welcomed by this :

BK560A lives again!

My BK560A rised from its grave! Not too bad for a 2-euro repair! (as much as the replacement part costed…)

 Posted by at 9:30 pm
Feb 052019

Got from USA an original Mortal Kombat PCB set complete with its sound board:

When I powered the board up the self-test reported two bad devices @U5 and U23:

P.O.S.T results

Then it booted into game but colors were completely wrong:

The two presumed bad devices are the color RAMs (Sony CXK5863 so 8K x 8-bit, pin-to-pin compatible with the more common 6264)

Color RAMs

Probing them revealed that pin 27 (Read/Write enable line) of both was stuck high:

This input signal is labeled on schematics as ‘-CRWR’ :

It comes from pin 22 of a 24 pin IC @U8:

EP600 ELPD color RAM control

The IC was faulty and needed to be replaced.At first glance I thought it was a simple PAL/GAL but actually it turned out to be an Altera E600 EPLD (so still a programmable logic device but something more complex with two clock inputs providing sequential logic too ).I had no chance to dump or reverse it but only to look for a donor board.Luckily I was sent another original Mortal Kombat PCB completely dead which I took the part from.Board was then fixed :

The bootleg I received for repair was the Yawdim one with no Midway logo on title screen :

It was fully playable but some backgrounds graphics were wrong :

As usual I did a visual inspection of PCB and found on solder side a scratch that apparently had severed a trace:

The involved area was the one populated by some 64K x 4-bit dynamic RAM so this was relevant to the kind of fault :

Checking with my multimeter confirmed the trace was really broken so I patched it, this restored background graphics and fixed the board:

Double repair accomplished.

 Posted by at 7:16 pm
Jan 272019

Got this game from a friend who took out the board out of a rusty cabinet.

The game needed badly a clean, so I dismounted everything before attempting to turn it on and washed everything carefully.

The game has two jamma connectors for player 1 ( Monitor A,left connector looking at the above picture) and player 2 ( Monitor B, right connector looking at the above picture). The connectors share same power lines, so you can connect wherever you want to play as P1 or P2.

Unfortunately board was not working correctly, it showed from a brief moment a garbled text about network connection and then nothing happened.

I decided to start dumping the program maskroms and found immediately one branded SEGA MPR-15539 which had some internal connection problems.

The other one , same type and capacity was dumped good.

After burning a new program rom on a 27c400 ( 4mbit maskrom pinout eprom), game booted perfectly although without any sound.

There was not even a noise coming from amplifier so I decided to desolder it and found immediately the problem

The amplifier was litterally blown

From Mame source I could get the right one to order, a TDA1518AQ, replaced it and finally the game had sound.

Unfortunately some tracks sounded bad with wrong samples.

There were other SEGA branded maskroms used as PCM samples, I read both of them ( 16mbit maskroms) and one of them , IC1, had A16 internally broken.

After buring a new 27c160 eprom with IC1 data, game sounded perfectly again.

No other problems were found

I will share the pinout of the board to play it on a different cabinet than the original one with a filter board ( the manual has only the schematics of the filter board unfortunately).

Game is Jamma as stated, so all power, video and audio pins are the same.

Board is using the following digital pins: on component side pin 22 is for shift up and pin 22 for shift down. On solder side, pin 22 is to select DJ Music or radio, pin 22 and pin 23 to move forward or backward the selection of musics. This is valid for both P1 and P2 connectors

There is an AMP connector in the middle used to drive lamps but I haven’t yet figured out the pinout and to me is not very interesting.

The most interesting part is the analogue controls which is handled but the daughter board 837-7636 which is in common with all Sega System 32 driving games.

I found a partial pinout on a japanese website and decided to expand it with Outrunner pinout

Jan 272019

Bought this game for personal collection but despite having declared working it had some graphic faults.

In particular the intro had blockly graphics and generally lacked shade of colours. The gameplay was in general fine at least for the first two levels.

I started to check maskroms and found LD13 and LD15 which handled the tiles of the intro. Maskroms were tested good on the programmer.

Data pins went to a 86s100 custom chip which usually is not very reliable ( see bottom right of the pic below)

At this point I decided to put a socket and change it with another one I took years ago from Poker Ladies.

Unfortunately the defect was still there but after another look underneath the board and I saw some borken traces which connected the custom to a 74ls273.

There were 3 traces severed but still the graphics were not fixed at all without a sign of improvement.

At this point I started to probe all the other ttl of the video board until I noticed a floating data pin of a 74ls273 which led again to the custom.

The trace interrupted at some point but I could not see any visible scratch until I noticed it near a silscreened text H1, not easy to find at all

After repairing this track the game was fully restored

Jan 222019

Got from Portugal five faulty G.I. Joe PCBs to repair (hopefully all of them…).I had a good start and could fix two of them.

The first board was in good shape with no missing parts:

It booted into game but colors were completely wrong and most of graphics missing:

I launched a MASK ROMs check which reported all the devices as good so the problem was elsewhere :

I remembered I had a similar issue in some of my past repairs, the culprit was the ‘053251’ custom ASIC which is present on my board too:

‘053251’ custom ASIC

This custom takes on input different layers of graphics as well as some priority bits and outputs 11 bit of palette index plus two shadow bits (info taken from MAME source)

‘053251’ schematics

I went to replace the IC with no further analysis:

Graphics were completely restored and board perfectly working with no other issue :

The second board was in good shape and complete too :

But it booted to a static pink screen :

Also in this case my suspicions fell on the ‘053251’ custom ASIC :

I replaced it, it was really faulty.Now the board tried to boot but failed the power on self-test:

The bad reported devices were all part of the sound circuit : @7C (SOUND ROM), 6C (SOUND RAM), @3E (‘054539’), @2D (62256 SRAM)

The sound CPU is a Z80B, probing it revealed floating address/data lines:

Luckily the CPU was aready socketed so I swapped in a good known one (a Z80E with clock up to 8MHz)

The board got now past of the self-test and booted into game but sound was absent:

Obvioulsy the culprit was always him, the ‘infamous’ ‘054986A’ hybrid module:

‘054986A’ hybrid audio module

As I usally do, I removed it :

‘054986A’ removed

and installed proper sockets:

1.778mm sockets installed

Then I serviced the audio module replacing the OP-AMP underneath :

and electroytic capacitors on top:

Eectrolytic capacitors replaced

Sound was fully restored and double repair accomplished.

 Posted by at 9:16 pm