22 Feb 2017

Sega Republic - Dubaï Mall

Not a repair log this time.
I've spent a week in Dubai during my holidays before Christmas and obviously stopped at Sega Republic in the Dubai Mall.
Here are few pics of some nice cabinets you can find there:

15 Feb 2017

Power Spikes - Video System 1991 (repair log)

The game works but with no sound.
Searching on the web, I found a guy who had 2 copies of the game with the same symptoms. He changed the Z80 (CPU used for audio) with no success.
That's what put me on the wrong way...

I saw that one of the data lines was stuck high and that the Z80 CPU had an active interrupt (NMI = Non Maskable Interrupt, which can be a cause or a consequence with multiple scenario that can lead to it, thus I wasn't too worried).
I dumped the audio ROM and it was identical to the MAME romset.
I pulled the audio RAM and it turned out to be good on my programmer.
Following the trace I found it was connected to one of the inputs of a LS174. I went further and I found it's also connected to one of the outputs of a LS374. The D signal was alive but the output (latched) was stuck high. Other latches of the same IC were working fine so I pulled it. And it was OK on my programmer... I pulled the LS174 cause sometimes even an input can be internally stuck high (quite rare though). It was OK too.
Without the LS374, LS174, audio RAM, audio and audio ROM installed (now all on sockets) the signal was still stuck high.
The only component left was the Z80. I pulled it, replaced it and the sound reappeared magically!

Game fixed.

 P.S. : I was a bit silly to think it couldn't be the Z80 because it wasn't the case twice for the same guy. There are so many parts on our precious board possibilities are almost infinite.

11 Feb 2017

Mystery Game (System 11) - Namco 1998 (repair log)

I got my hands on a unidentified Namco System 11 game (hardware based on the Sony Playstation).
By swapping the boards with another known working system 11 game, I could determined that the issue lied in the CPU board.

I quickly found a missing signal on a LS08 with my logic probe. Following the trace I saw it was connected to the main CPU.
In fact the board took a shock (poor storage) and around 30 pins were desoldered on the main CPU:

I fixed it using hot air and the game then booted:

Drum roll: it's Kosodate Quiz My Angel 3... Awe... some...

First I already own this game, second it's a quiz game in Japanese where you play a girl from birth to age 25 with the ultimate goal of getting married.

1 Feb 2017

Super Nintendo hacked pad for Virtual Boy

Virtual Boy controllers are quite uncommon so I had to find an alternative solution.

The NES, SNES and Virtual Boy share the same serial communication protocol with few differences :
- NES controllers only send 1 byte of data (8 bits) whereas SNES and Virtual Boy controllers send 2 bytes of data (16 bits) due to the fact they have more buttons
- Button mapping is slightly different between the SNES and the VB (well not really if you look at the physical position of buttons, the left D-pad of the VB corresponding to buttons B/A/Y/X on the SNES pad)

In order to build my adapter I pulled a SNES controller port from a dead unit and for the plug on the VB side I used a NES one I cut a little to make it fit in the VB port:

The adapter also includes the PSU plug of the SNES (with the Graëtz bridge and the 7805).

The data sent by the SNES are in the following order:
B-Y-Select-Start-Up-Down-Left-Right-A-X-L-R = 12 bits, the remaining 4 bits are always high
For the VB:
Down(R)-Left(R)-Select-Start-Up(L)-Down(L)-Left(L)-Righ(L)-Right(R)-Up(R)-L-R-B-A = 14 bits, bit 15 being the low battery indicator and the last one being always high

The right D-pad on the VB having no real use as far as I know, I decided to reroute Down(R) and Right(R) to buttons A and B.

[EDIT] Seemingly 3 games can be problematic:
- 3D Tetris: the right D-pad is used for block rotation, so the lack of Down(R) and Right(R) means you have to press Up(R) or Left(R) 3 times instead of 1 to obtain the desired orientation.
- Red Alarm: in this game buttons A and B have the same function as Select and Start thus you can play the game with an unmodified SNES pad.
- Teleroboxer: unfortunately this game needs every single button. One solution is to add 2 buttons to a SNES pad.

Back to the mod.
Two 4021 are daisy chained to serialise data. What have to be done:
- reroute button A from pin 1 of the right 4021 to pin 5 (of the same 4021)
- reroute button B from pin 1 of the left 4021 to pin 4 of the right 4021
- pull pin 1 of both 4021 to +5V, if not the console will see Down(R) and Right(R) as pressed

Pin 1 of the left 4021 and pins 1/4/5 of the right 4021 have to be lifted (be careful they are mounted head to tail).

Then the following connections have to be made:
- pad 1 of the right 4021 has to be connected to pin 5 (lifted) of the same 4021
- pad 1 of the left 4021 has to be connected to pin 4 (lifted) of the right 4021
- pin 1 (lifted) of both 4021 have to be connected to +5V

The pads can be difficult to access so you can solder on the resistors:

Wiring done: