Thursday, 25 May 2023

Nintendo Vs Super Mario Bros pcb repair

I got this Nintendo Vs board in a trade, someone had taken a heatgun to it at some point and done some damage to the board along with taking off a bunch of TTL, '104' capacitors and the eproms.

Once repopulated and fixing a blow out via caused by the heatgun damage I powered it on but the game did not run, just a solid green screen.

It's not shown in the picture but the top part of the board had a standard cpu installed, with Vs Super Mario Bros you need a dummy cpu in this socket for the game to be run.

Thankfully you can get away without having a dummy cpu and just bridge two pins on the cpu socket instead.


This got the game running but with some strange issues such as bad graphics, messed up scores, phantom coin inputs and other times it would run fine for 10-15 minutes.

I was worried it was going to be an issue with trace(s) caused by the heatgun damage and I was going to have to check every connection in the area with a multi meter. Before all that I went through and checked all the ram and TTL chips first as I already knew the cpu / ppu were fine from trying them on another board.

The trusty boardmaster 4000 flagged a 74LS157 @ 5J as having a stuck low output on pin 7.

This was confirmed out of circuit, a new 157 fitted and it is now fully working.

Monday, 28 November 2022

Early Donkey Kong with different colour palette

I picked up a couple of Donkey Kong 4 stack pcbs for repair, these are the type typically found in cocktail cabinets. The repairs weren't all that interesting typical stuff for DK boards (bad sockets, bad 8257 resulting in missing sprites, bad fujitsu TTL).

What I did find interesting is that one of them is a very rare and early TKG1 revision Japanese version that even predates the Radarscope boards that were converted to Donkey Kong.

I suspect there was a small run of boards made to put out on location to see how well the game performed before Nintendo invested too much money into the game. They wouldn't want another Radarscope on their hands where they got stuck with boards and cabinets they couldn't sell.

It has the original version of the game which has the levels in a different order, the ladder cheat and allows you to save long names during high score saving, the colour palette is also a little different. It is also a bit more standardized than later versions as it outputs regular video that isn't inverted and there's an on board amplifier for the audio which is usually on a pcb attached to the monitor.

The proms have hand written numbers, someone could argue that the proms are simply bad or from another game but as another board was found with the exact same 'off' colours (onecircuit on YT has a video) I think that confirms that originally Donkey Kong was released with these colours and Nintendo changed them later before the TKG2 boards were shipped.

It's not too obvious from these pictures but a lot of the colours are lighter shades, especially Donkey Kong who is a much darker brown in all other versions of the game.


I could only find two other examples of this board online but both have this Kyodo sticker on the sound board.

I don't think there's any way to easily replicate the colours in mame as the proms aren't actually read by mame but for the sake of archiving the proms I've uploaded the dumps here.

Sunday, 16 October 2022

2 years went fast..

Not much to report, I haven't got many repairs done in the last 2 years. Starting to get back into it now, I might post some repair logs... I see the only comments I got was due to broken links, I have switched to using a google drive so hopefully that's more reliable.

For some reason the large collection of pdf manuals and schematics that were hosted at are gone and it looks like the majority of them haven't made it to either. So for now at least I've stuck what I have on my google drive.


The download link is in the top right hand corner, I don't have all of them, not even close. What's there are mostly ones I needed to use myself, there are a couple that I don't believe were ever hosted any where though, and some I cleaned up myself so have a browse. 😇

Tuesday, 1 December 2020

Sega Mega-Tech pcb repair

I wasn't planning to log this one which is why there aren't any before pictures but there turned out to be quite a lot wrong with it.

First thing I had to do was clean it as it was filthy, powering it on just gave a black screen. On closer inspection all the sockets needed to be replaced as they were badly corroded.

This got the top screen to work but the games weren't listed or running. This is usually down to bad Sony CXD1095Q chips, I ordered a couple from aliexpress and replaced both of them. This fixed the games not loading / displaying issue but it also revealed new problems.

The graphics were missing and just displaying noise and blocks, this was down to the two video drams that again I had to order. I found some nice new old stock M5M4C264AL-12 on ebay which are the same type and speed rating as the originals.

Now the games were playable but I noticed certain sounds were missing, 'Sega' and the ring collecting sound in Sonic were missing along with a lot of other sounds. This was due to a bad YM2612 which thankfully I had some spares on hand and didn't have to wait a third time.

I also noticed the red output on both screens was quite weak, especially on the main screen. This was due to faulty and leaky transistors, replacing them fixed the problem.

That looked to be everything but I found one final fault, about half the dip switches weren't working. Replacing them with new ones fixed the problem and the board is finally 100%.


Repair complete.

Friday, 31 July 2020

Atomiswave fan upgrade

I picked up this Atomiswave motherboard in a recent trade, I've been after one for years as I had a Metal Slug 6 cartridge (bootleg) that I haven't been able to play.

It worked great but the original fan was screaming and very distracting.

I used to get some nice silent red 40mm fans from ebay for CPS2 motherboards, this would have been perfect as it would have matched the Atomiswave case. They're no longer available though so I had to find something else, I ended up settling with this one.

Noiseblocker Black Silent XM-2, it is very quiet still moves some air and most importantly was cheap (£7 posted). There is an MX-1 version too which is even quieter but it doesn't move much air so I wouldn't recommend it. The XM-2 is very quiet to the point you won't hear it during game play or if you have the motherboard installed in a cabinet.

As with most things it's not just a simple case of screwing and plugging it in, you'll need to take the original cable and splice it with the new one as the connectors are different.

The third yellow wire on the original fan feeds ground back into the board from the black wire, since the new fans white wire is for sens this won't work and will cause your board to lose sound.

What you have to do is also bridge the second and third pins on the motherboard with a solder blob (this is also covered on arcade-project forums).

That's it, one quiet Atomiswave.

Now to try and track down some more cartridges.

Wednesday, 15 July 2020

Blood Bros pcb repair

This one was a real headache, it was completely dead with nothing on screen and no sound. After going down many different rabbit holes I checked the previous work that had been done which is the first thing I should have done.

The 74LS04 at U071 had been socketed after removing the socket I could see quite a bit of damage, including the trace from pin 11 to 13 being broken.

After patching the bad trace it would finally power up and display something, but there was no sync and no sound.

The sync issue was fixed by reflowing the SEI0200 custom chip.

I also noticed the LS04 at U053 had been replaced with a HC04, it may have been okay to use a cmos chip here but I replaced it with the proper LS04.

The only issue remaining was the lack of sound and the coin inputs not working. These are linked as the Z80 wasn't running which is the sound cpu and for whatever reason seibu also used the Z80 for the coin inputs.

The main sound eprom was bad but after replacing it the cpu still wasn't running.

I next desoldered the Z80 and fitted a socket so I could hook up the Fluke 9010a and also test the cpu, the cpu was fine.

The fluke was reporting 'Data bits 0 tied' but tied to what? usually it will say ground, 5v or another line. According to a cource notes fluke pdf document I found online some combinations of shorts or when it's multiple lines the fluke cannot always identify them and cannot give you any more details.

Checking with a multi meter I wasn't able to find any shorted lines, I lifted the D0 pins on various chips and wasn't able to find the issue. I just had to start pulling and socketing stuff directly connected to the data bus after pulling the M6295 pcm voice chip the bus test on the fluke passed.

I have seen these fail quite a bit on CPS1 motherboards but never like this where it kills the data bus. It looks like they are getting very hard to find now too. A couple of sellers in China have them but they look remarked so who knows what they really are. Luckily I had a couple of working ones in my spares drawer pulled from scrap boards.

M6295 fitted and the sound and coin inputs are now all working.

Repair complete.

Friday, 5 June 2020

Super Off Road (Ivan Ironman Stewart's) repair and notes

A Super Off Road board came in for repair, the first thing I had to do was make up a basic power and video harness to test it. It was completely dead, a look over the board revealed it was two non working pcbs (different serials) from different sets connected together with a big X in marker on each.

It had also been stored in a damp place so it had some corrosion and every socket on the board needed to be replaced, a lot were single wipe rubbish that would have to be replaced anyway. Apparently Atari made these boards or at least assembled them, I'm not sure if this is true but it would explain the sockets.

This board uses a LOT of PLDs (Programmable logic devices), all secured and none have been dumped or replicated from what I can tell. This was my main concern as one bad PLD and I wouldn't be able to get it going.

Top Board PLDs

Location   Device          Name
U9            PAL20L10    38-01
U42          PAL16R6      37-01
U43          PAL16R4      36-01
U44          PAL16L8      35-01 - Bruteforced, jed file available at
U54          PAL16L8      41-02 - Bruteforced, jed file available at
U55          PAL18P8      39-01
U68          PAL18P8      40-01
U74          PAL16R6      19-01
U75          PAL16R6      18-01
U82          PAL16R6      21-01

Bottom Board PLDs

Location   Device          Name
U16          PAL16L8      03-01 - Bruteforced, jed file available at
U17          PAL16L8      04-01
U19          PAL20X10   06-01
U21          PAL20X10   05-01
U22          20X8                       Programmed to behave like a 74LS461 which can be used instead.
U23          20X8                       Programmed to behave like a 74LS461 which can be used instead.
U26          PAL16R8     10-01
U27          PAL16R8     11-01
U40          PAL16R8     09-01
U42          PAL16R4     08-01
U43          PAL16R8     07-01
U45          20X8                       Programmed to behave like a 74LS461 which can be used instead.
U56          20X8                       Programmed to behave like a 74LS461 which can be used instead.
U60          20X8                       Programmed to behave like a 74LS461 which can be used instead.
U85          PAL16L8     02-22
U96          PAL16L8     01-01 - Bruteforced, jed file available at

I bruteforced, created jed files and tested the four PALs I could, this still leaves a lot that cannot currently be replaced but it's a start.

On to the repair.

The board would boot from cold and display the title screen but it was cut up into pieces and then would crash. Trying to boot the board again or pressing the reset button would result in vertical lines and the slave cpu resetting.

When replacing sockets I tested every chip I removed and found two bad 6116 rams and two bad 41464 rams which I replaced. I also found that R1 resistor next to the CPU socket U1 on the bottom board was only soldered on one end and the other end was just touching the pad, it must have been like this since it left the factory, so I soldered that on properly.

It still behaved the same though, next I found a bad 74LS374 @ U18 on the bottom pcb, replacing this fixed both the graphics and crashing issue.

I'll also replace the original battery from 1989 with a new CR2032 coin battery and holder.

The sound and digital controls all work but I haven't been able to test the steering as I don't have the 'opto coupler' board but hopefully this repair is complete.