xodaraP
Legendary
Hi guys,
This is a repair log I've been working with on another site but I thought I'd add it here too for anyone interested
Big thank you to Womble for his repair logs and his help so far with
this one. It's not completely working yet but I'll update it as I go.
I have a Sega System 16A boardset here I bought a long time ago as not
working and was thinking of using it for parts. The board was completely
dead, very dusty and I put it aside in the potential scrap pile as it
seemed like it was never going to live again. It was marked as a Shinobi
and I was thinking if I ever came across a semi working boardset I
would have the parts needed to fix it. One day I was reorganising boards
and looked in the scrap pile and found it again, looked it over and
decided since it was actually complete with all the cables it would be
worth the time to at least attempt a repair before resigning it to the
scrap pile. I also found it was originally a Fantasy Zone after I
cleaned up the original label on the board so this repair will not only
be a repair of the board but also deconverting it back to it's original
form as a Fantasy Zone
So to start with I have this boardset, complete but not working, currently I get this when I power it on:
https://www.dropbox.com/s/gayowfjqrhz2h27/File 5-12-16, 13 02 51.jpeg
It is complete however I had pinched a chip from it to repair another
board (LS00) - I did a pretty terrible job removing the chip and
accidentally damaged a couple of the pads. I'm not proud of this, but I
have now fixed it. Originally I replaced it with a socket but I removed
that and replaced the chip properly, I have also run wires to repair the
traces so the board is then complete. The solder job works, I agree
it's ugly but as long as it works I'm not concerned about how it looks
since it's hidden between the board stack
https://www.dropbox.com/s/gayowfjqrhz2h27/File 5-12-16, 13 02 51.jpeg
I've made it look as clean as I possibly can, but feel it needs to be
shown anyway. On top of that when I was moving some stuff around on my
repair area I accidentally dropped a 1kg roll of solder onto the board
and completely smashed a HC04, luckily Jaycar sells these so I bought
one and replaced it. Lastly I also accidentally pulled a pad off when I
was removing and testing some of the Fujitsu TTLs on the board so I ran a
wire to repair this as well.
https://www.dropbox.com/s/401m702rfsty5pp/File 5-12-16, 13 04 59.jpeg?dl=0
After replacing these I have no change in the picture when I boot the
board up. I noticed that the clock signal is also missing from the CPU
so I traced this back through quite a few Fujitsu ICs which I also
removed and tested off board as I went, and replaced the 16Mhz crystal
on the board which may have also been damaged when the roll of solder
dropped onto the board and damaged the HC04 next to the crystal. I also
replaced all of the power filtering caps on the top board as a couple of
these were swollen and one had been half wrenched off the board. These
are 470uf 16V caps and easy to find.
I now have the same image as above when the board is booted, however I
now have a clicking sound and the graphics flicker - the board has life!
The CPU is trying to boot the game! However all I'm getting is garbage.
I'm completely stuck here, I try to run the top board only but the CPU
doesn't start and run anything without both boards being present. All
the logic on the top board checks out so I start to check and make notes
about the bottom board. I do notice that a few of the ICs have floating
pins, but this could be because they're not getting data due to the
system not actually starting up.
After talking with Womble about possible issues, he suggested that I
check the data in the main CPU ROMs - I had already dumped the full set
of what was on the board and it was identified as Shinobi, System 16A,
set 6, Unprotected which seemed fine. The ROMs at positions 26 and 43
were specific to that set, 25 and 42 (the other main CPU code ROMs)
identified as being part of set 6 as well as being the same as sets 2
and 3 from the System 16B version of the game. I had no idea if set 6 is
usable on original hardware, but decided to try the decrypted set from
Porchy's JAMMArcade.net decrypted Sega section. These were not listed as
having been tested on original hardware, however they did not come up
in ROMident so they were not the same code as the set 6 that was
currently on the board.
I now have this:
Success! The board is running the attract mode! The sound appears to be
perfect but the video signal is a complete mess. Thanks to Womble's
repair log here http://www.retrocomputermuseum.co.uk...p?topic=5288.0
- I know that the composite sync from the board is provided by an LS86
at IC57 via Pin 6 on the bottom board which takes the VSYNC signal from
the PAL at IC35 through Pin 5 and HSYNC is provided from PALs at IC58
and IC66 through Pin 4 which are fed by a group of LS669s -the LS86 is a
Fujitsu and so I'll replace it anyway, but looking at it with a logic
probe it actually appears to be the most likely culprit. Pins 4 and 5
are both showing high and low pulses where pin 6 is showing low pulses
only and is pulsing much faster than pins 4 and 5.
I have now replaced the LS86 and there is no change, however Womble has
suggested that this is a video ground issue, not a sync issue so it may
be that an extra wire needs to be connected up on the harness.
I have now installed a video ground wire, the system 16 harness I have
had the video ground connection on the JAMMA side (pin 15) unconnected, I
installed a small jumper wire to the ground on the connector and it
came up, I now have this:
Still some work to do, but it's a lot further than it was when I
started. Looks like the background graphics are at fault, the sprites
appear to be ok. The background graphics are at IC93, 94 and 95 on the
top board and I have pulled these, they have tested ok. The on board
tests also suggest that the RAM ICs are ok as well, so it's likely to be
the logic at fault. I'll stab around with the logic probe in the area
and see what I come up with.
UPDATE: So I stabbed around with the logic probe and followed the fault
from the ROM at IC93 back to the LS273 driving it which had a floating
pin at pin 7. I swapped out the LS273 even though it tested fine off
board because it's a Fujitsu and because I'd already removed it but it
was working fine by the look of it. I then followed the fault back to
IC80. A custom Sega chip silkscreened 315-5025 on the board but mine was
populated with a 20 pin 315-5155 (the 315-5025 is 18 pin), the chips
are the same and there are 6 of them in this area of the board, 2 for
each of the graphics ROMs. Pin 3 on IC80 is floating (not showing any
data at all) meaning it needs replacing. I will have to chase myself
down one of these chips and replace it.
Until then, to be continued.
Some notes for anyone with similar issues trying to repair your board:
The MAME set 6 for Shinobi on 16A hardware does NOT have the correct
maincpu code - use the set from JAMMArcade.net. HOWEVER the naming on
the ROMs in the JAMMArcade set is incorrect at the time of writing. The
names were marked as IC27 which is a RAM chip, this ROM actually goes
into position 26. The other ROM is marked as IC42 which is a ROM, but
this is the incorrect ROM needing to be replaced. IC42 from the
JAMMArcade.net set needs to go into position 43. The other ROMs from set
6 seem to be ok including the other 2 maincpu code ROMs.
The PAL on the top board at position 69 for Shinobi is 315-5244 and is
VITAL for the board to function. This PAL is used for assigning RAM and
ROM locations and the board will not boot without it. This PAL supports
the larger ROM sizes and will read 27c512 ROMs for the maincpu code. It
can be burnt to a GAL16V8 and is available from JAMMArcade.net. Fantasy
Zone uses 27c256 for everything and uses a slightly different PAL which
has not yet been dumped or converted for use with a GAL, it is PAL
315-5150 and can be replaced with the 315-5244 GAL16V8 however some
modifications need to be made to your ROMs to make them fit onto a
27c512. This information is available elsewhere.
This is a repair log I've been working with on another site but I thought I'd add it here too for anyone interested
Big thank you to Womble for his repair logs and his help so far with
this one. It's not completely working yet but I'll update it as I go.
I have a Sega System 16A boardset here I bought a long time ago as not
working and was thinking of using it for parts. The board was completely
dead, very dusty and I put it aside in the potential scrap pile as it
seemed like it was never going to live again. It was marked as a Shinobi
and I was thinking if I ever came across a semi working boardset I
would have the parts needed to fix it. One day I was reorganising boards
and looked in the scrap pile and found it again, looked it over and
decided since it was actually complete with all the cables it would be
worth the time to at least attempt a repair before resigning it to the
scrap pile. I also found it was originally a Fantasy Zone after I
cleaned up the original label on the board so this repair will not only
be a repair of the board but also deconverting it back to it's original
form as a Fantasy Zone
So to start with I have this boardset, complete but not working, currently I get this when I power it on:
https://www.dropbox.com/s/gayowfjqrhz2h27/File 5-12-16, 13 02 51.jpeg
It is complete however I had pinched a chip from it to repair another
board (LS00) - I did a pretty terrible job removing the chip and
accidentally damaged a couple of the pads. I'm not proud of this, but I
have now fixed it. Originally I replaced it with a socket but I removed
that and replaced the chip properly, I have also run wires to repair the
traces so the board is then complete. The solder job works, I agree
it's ugly but as long as it works I'm not concerned about how it looks
since it's hidden between the board stack
https://www.dropbox.com/s/gayowfjqrhz2h27/File 5-12-16, 13 02 51.jpeg
I've made it look as clean as I possibly can, but feel it needs to be
shown anyway. On top of that when I was moving some stuff around on my
repair area I accidentally dropped a 1kg roll of solder onto the board
and completely smashed a HC04, luckily Jaycar sells these so I bought
one and replaced it. Lastly I also accidentally pulled a pad off when I
was removing and testing some of the Fujitsu TTLs on the board so I ran a
wire to repair this as well.
https://www.dropbox.com/s/401m702rfsty5pp/File 5-12-16, 13 04 59.jpeg?dl=0
After replacing these I have no change in the picture when I boot the
board up. I noticed that the clock signal is also missing from the CPU
so I traced this back through quite a few Fujitsu ICs which I also
removed and tested off board as I went, and replaced the 16Mhz crystal
on the board which may have also been damaged when the roll of solder
dropped onto the board and damaged the HC04 next to the crystal. I also
replaced all of the power filtering caps on the top board as a couple of
these were swollen and one had been half wrenched off the board. These
are 470uf 16V caps and easy to find.
I now have the same image as above when the board is booted, however I
now have a clicking sound and the graphics flicker - the board has life!
The CPU is trying to boot the game! However all I'm getting is garbage.
I'm completely stuck here, I try to run the top board only but the CPU
doesn't start and run anything without both boards being present. All
the logic on the top board checks out so I start to check and make notes
about the bottom board. I do notice that a few of the ICs have floating
pins, but this could be because they're not getting data due to the
system not actually starting up.
After talking with Womble about possible issues, he suggested that I
check the data in the main CPU ROMs - I had already dumped the full set
of what was on the board and it was identified as Shinobi, System 16A,
set 6, Unprotected which seemed fine. The ROMs at positions 26 and 43
were specific to that set, 25 and 42 (the other main CPU code ROMs)
identified as being part of set 6 as well as being the same as sets 2
and 3 from the System 16B version of the game. I had no idea if set 6 is
usable on original hardware, but decided to try the decrypted set from
Porchy's JAMMArcade.net decrypted Sega section. These were not listed as
having been tested on original hardware, however they did not come up
in ROMident so they were not the same code as the set 6 that was
currently on the board.
I now have this:
Success! The board is running the attract mode! The sound appears to be
perfect but the video signal is a complete mess. Thanks to Womble's
repair log here http://www.retrocomputermuseum.co.uk...p?topic=5288.0
- I know that the composite sync from the board is provided by an LS86
at IC57 via Pin 6 on the bottom board which takes the VSYNC signal from
the PAL at IC35 through Pin 5 and HSYNC is provided from PALs at IC58
and IC66 through Pin 4 which are fed by a group of LS669s -the LS86 is a
Fujitsu and so I'll replace it anyway, but looking at it with a logic
probe it actually appears to be the most likely culprit. Pins 4 and 5
are both showing high and low pulses where pin 6 is showing low pulses
only and is pulsing much faster than pins 4 and 5.
I have now replaced the LS86 and there is no change, however Womble has
suggested that this is a video ground issue, not a sync issue so it may
be that an extra wire needs to be connected up on the harness.
I have now installed a video ground wire, the system 16 harness I have
had the video ground connection on the JAMMA side (pin 15) unconnected, I
installed a small jumper wire to the ground on the connector and it
came up, I now have this:
Still some work to do, but it's a lot further than it was when I
started. Looks like the background graphics are at fault, the sprites
appear to be ok. The background graphics are at IC93, 94 and 95 on the
top board and I have pulled these, they have tested ok. The on board
tests also suggest that the RAM ICs are ok as well, so it's likely to be
the logic at fault. I'll stab around with the logic probe in the area
and see what I come up with.
UPDATE: So I stabbed around with the logic probe and followed the fault
from the ROM at IC93 back to the LS273 driving it which had a floating
pin at pin 7. I swapped out the LS273 even though it tested fine off
board because it's a Fujitsu and because I'd already removed it but it
was working fine by the look of it. I then followed the fault back to
IC80. A custom Sega chip silkscreened 315-5025 on the board but mine was
populated with a 20 pin 315-5155 (the 315-5025 is 18 pin), the chips
are the same and there are 6 of them in this area of the board, 2 for
each of the graphics ROMs. Pin 3 on IC80 is floating (not showing any
data at all) meaning it needs replacing. I will have to chase myself
down one of these chips and replace it.
Until then, to be continued.
Some notes for anyone with similar issues trying to repair your board:
The MAME set 6 for Shinobi on 16A hardware does NOT have the correct
maincpu code - use the set from JAMMArcade.net. HOWEVER the naming on
the ROMs in the JAMMArcade set is incorrect at the time of writing. The
names were marked as IC27 which is a RAM chip, this ROM actually goes
into position 26. The other ROM is marked as IC42 which is a ROM, but
this is the incorrect ROM needing to be replaced. IC42 from the
JAMMArcade.net set needs to go into position 43. The other ROMs from set
6 seem to be ok including the other 2 maincpu code ROMs.
The PAL on the top board at position 69 for Shinobi is 315-5244 and is
VITAL for the board to function. This PAL is used for assigning RAM and
ROM locations and the board will not boot without it. This PAL supports
the larger ROM sizes and will read 27c512 ROMs for the maincpu code. It
can be burnt to a GAL16V8 and is available from JAMMArcade.net. Fantasy
Zone uses 27c256 for everything and uses a slightly different PAL which
has not yet been dumped or converted for use with a GAL, it is PAL
315-5150 and can be replaced with the 315-5244 GAL16V8 however some
modifications need to be made to your ROMs to make them fit onto a
27c512. This information is available elsewhere.
