PCB Repair Logs – Page 110

CAPCOM Commando repair log #3

Apr 162011

Whilst soak testing the Commando board I had recently repaired for someone 2 bad things happened.
First the music disappeared
Second the pictures vanished.

I checked the RGB outputs with the scope and the signals looked a little weak but should be fine. I checked the SYNC signal and this was stuck HIGH.
Followed the signal back from the edge connector to a 74LS138 IC @ location 5L and the outputs on pins 3 and 4 were both stuck at just under 1v according to the scope.

The inputs to the IC were fine. I piggy backed a working LS138 and the picture came back normal.

The sound fault was fairly simple too.
The first thing I did was check the RESET and HALT lines on the Z80 CPU. These were fine and not stopping operation. Next I checked the clock signals on the Z80 and the Yamaha 2203’s. The Z80 was fine but there was nothing on the 2203’s.
The clock signal for these 2 come from a nearby 74LS74 IC who’s outputs were obviously dead but all the inputs were “as schematic”. I pulled this and testing it, it failed and a replacement has restored the sound fully.
Finally this is in good health once again and hopefully it will stay that way for a while longer, although with the amount of Fujitsu TTL IC’s on the board I very much doubt that but it is 26 years old now!

CAPCOM Commando repair log #2

PCB Repair Logs, Repair Logs Comments Off

Apr 102011

Got sent a Commando board for repair a while back and have finally got around to giving it a test.
I spent a little time going over the board doing the visual inspection prior to booting up. On the bottom board a 74LS245 had been removed and was just sitting in the board loose but most of the traces going to the IC had also been lifted, I suspect during a bodged removal.

I refitted the IC (after testing it), cut the lifted tracks off the board and ran some kynar to the relevant points.

After spending a bit of time making another adapter up to run the game I saw the graphics fault.
All the sprites were barely visible and sometimes were gone completely.
There was also some corruption on one of the screens.

The manual for this game is online and in the back of this are the schematics for the game. Nice. Also, as its CAPCOM, the schematics are nicely split and labelled up in sections.
After seeing the fault I was convinced that RAM was to blame. I scoped a section of 4 x 2114 RAM chips on the lower board. All had good looking address lines but the data lines looked very odd like they were struggling to drive the lines low. I interrupted a couple of pins and graphics started to come back so I pulled it and testing it, FAIL. One replacing the RAM only half the sprites came back with jail-bars running through them. I ended up pulling 3 in total which was quite lucky as I ran out of RAM replacements.
All graphics are now returned to normal.

As for the corruption on the screen. Thanks to MAME I was able to determine this to be an unused part of the screen that my RGB to VGA converter decided to throw in there to confuse me. In a cabinet the screen would be adjusted to remove these borders.

Double Dragon repair log #4

PCB Repair Logs, Repair Logs Comments Off

Mar 122011

The sound circuits on these DD boards really were bad!
Another one with dodgy sound, music was intermittent and half the SFX were really really low.
Did a full capacitor replace on this board for the sound, half the caps had started to leak. This brought back the music.
The sound effects were a little confusing. Like the last board I could see the data going into the M5205 chips but on IC80 there was nothing coming out of pin 10. I replaced the chip but this time there was no change. The board has had the op amps replaced at some point and there was a little rework done on them. Pin 5 (input) of the op amp at IC97 was shorted to pins 6 and 7 (output) effectively bypassing the op amp altogether, this explained why the sound was really low. Removing the short fixed the sound.
Another one destined for eBay.

Rainbow Islands repair log #1

PCB Repair Logs, Repair Logs Comments Off

Mar 052011

Got hold of a very poorly Rainbow Islands Extra board.
This board had some repair work done previously. There are a few burned out tracks and some dodgy looking wire links.

On power up it just showed a black screen, time to break out the Fluke 9010a.
The Fluke showed I had active interrupt lines. The 68000 CPU has 3 interrupts, IPL0, IPL1 and IPL2. IPL1 and 2 are tied to +5v (they are active low therefore unused here). IPL2 was stuck low.
I traced the line to a nearby 74ALS74 chip and then back to a 74LS273 (IC32). Regardless what the inputs were doing, the outputs were all stuck low. I replaced this and tested, still a blank screen. I check the 74ALS74 again and although all the inputs were now present and correct, its output to IPL2 was still stuck low. I piggybacked a new IC onto it and the game came to life. Replaced the IC.

Now the game booted up I could see that the colours of the character sprites were wrong, mainly black.

This caused me a lot of headache as the colours came good when I put the logic probe on different pins of different IC’s. I determined which IC was partially responsible for sprite colours by interrupting lines on various logic chips. I found that if I toggled pin 9 of IC35 (74LS157) the colours came good more regular. I removed this chip and tested it in my logic tester, it passed but if I put it on a loop it would intermittently fail. Replaced this and the colours were now back to normal.

There is also a problem when the sprites are facing left where it shows a broken character but still facing right. The pictures below shows the broken up boss sprite and broken rainbows when fired to the left.

This ended up being a short on pin 15 @ IC38 (74LS153). This pin was shorted very slightly to an exposed track running next to it. Removing this short fixed the graphics completely and the game is now 100%.
Very pleased.