Nov 082017

Received this quite rare Hard Puncher (on Konami Twin16 hardware) for repair:

PCB failed all the time the self-test showing a bad device @10G on CPU board and then resetted, this in an endless loop:

The involved device is a Mitsubishi M5M5165 (8K x 8-bit static RAM) compatible with 6264 :

Its data bus showed no activity:

The chip failed the out-of-circuit test each time in a different address location:

With a good RAM the board successfully booted into game:

But speeches were missing, here is a comparison with a MAME recording:

I quickly pinpointed this fault in a missing 640KHz ceramic resonator, this is for generating clock to the uPD7759 ADPCM speech synthesizer IC.I took a spare from a dead Sega System16B motherboard:

Job done.

 Posted by at 9:13 am
Nov 042017

The second of Alex’s board for repair is Donkey Kong Junior.
This board booted to static screen that wouldn’t sync. This ended up being a broken 30k pot on the H-Sync.

As I don’t currently have and 30k or 50k pots I opted to temporarily fit an 18k resistor to the video PCB to let me move on.

The game now booted to a static screen of garbage but when touching the Z80 CPU it booted. The socket looked old and a bit crusty so I replaced it.
The game now played but there was a lot of garbage still on the screen.

I hooked up the Fluke 9010 and did some basic reads and writes to the video RAM that sits between address 0x7400 – 0x77ff.

Clearly bits 0 and 1 were stuck low.
Using the schematics I started checking at a 74LS245 at location 6A which buffers the databus.

Straight away I found although the signals were going to the chip the outputs on DAT0 and DAT1 were floating.
As the outputs were floating I tested by piggybacking a good 245 chip and everything came up good.

Replace the 245 and fitted a new one.

Tested the game and all sound and controls work too.

Another one fixed.

 Posted by at 2:52 pm
Nov 042017

Got a Donkey Kong 3 to repair from my good friend Alex @ Nintendo Arcade
On boot up I got a static green of garbage

I checked the voltages at the far end of the PCB and saw they were quite low compared to what my PSU was set at. Adjusting the voltage to a bit higher brought the PCB voltage back up to a good point but made no difference to the fault.
Next job was to check the program ROM’s

Found 7C, 7D & 7E had suffered from bit rot. I erased these and reprogrammed them with the correct data from the MAME set.
It didn’t make any difference at all so I started looking at the Z80 CPU signals and this is where I got a little confused.

At first glance and without really engaging brain too much I automatically assumed that there were two Z80 CPU’s side by side but checking the voltages and the signals left me scratching my head as they were not what I had thought. On closer inspection I realised that the one on the left is actually a Z80 DMA chip. I’ve not seen or even heard of these before so it something new to me.
I took to the internet and downloaded the datasheet for the DMA chip and also the schematics for this PCB too.
The CPU and the DMA chip are tied to the same busses and the DMA chip didn’t appear to be releasing the data bus to allow the CPU to do its thing.
I checked all the incoming signals to the DMA chip but found nothing odd so at this point I figured I needed to buy a new DMA chip so ordered one from eBay.
To help confirm my diagnosis I pulled the DMA chip and powered up the game. Sure enough it booted fine and ran but was missing the sprites which is to be expected really.

A couple of days later I received the new chip and now we get this

All sound and controls working too so that’s this one sorted.
There appears to be some skewing at points in the pictures above. This is due to my supergun setup and not a PCB fault.

 Posted by at 12:41 pm
Nov 012017

Got this Super Burger Time PCB (by Data East) for a repair:

Board played fine except for the fact the BLUE color was totally missing hence the yellowish screen:

Palette RAMs are two 2K x 8-bit devices, both were Sony CXK5814 (6116 compatible) which are well known to be unreliable:

Probing them revealed that the one @9K had all data lines stuck low while address were happily toggling (on left of the below picture) 

The IC was most likely bad so I removed it:

The out-of-circuit confirmed it:

Replacing it fixed the board completely.

 Posted by at 6:36 pm
Oct 252017

Received this faulty Capcom CPS1 A-BOARD, actually the ‘DASH’ revision with a 12MHz oscillator in place of 10MHz one:

Board played with good graphics but sound samples were completely wrong and randomly played, here’s a video made with a Captain Commando B-BOARD/C-BOARD:

PCM samples are played by an OKI ‘MSM6295’ IC which addresses two 1Mbit devices (located on B-BOARD) and reads back their data.Judging from flux residuals the IC was previously reworked on this board as well as other components of audio circuit:

For first I tested all traces to/from it, all was OK.So, most likely, the IC was bad, no wonder since it’s a prone to failure part.I removed it:

Cleaned the PCB from flux and installed the spare:

Time to test:

Board fixed.End of job.

 Posted by at 6:49 pm