NeoGeo MVS MV1ACH - Color RAM error on boot

nvana

n00b
Joined
Jan 20, 2022
Posts
3
Hello,

I'm trying to repair a MV1ACH board (is it a chinese version?) board giving me a Color RAM error on boot.

I've exchanged the 2 Palette RAM chips with a donor board, no difference.

I've exchanged the Neo-Go buffer IC leading upto the Palette RAMs, no difference.

Do you specialists have any tips of what else could be broken ? I'm out of clues.

Did anyone of you find some digital schematics of the NeoGeo MVS boards, currently I only have some fuzzy scanned PDFs.

It would make debugging the NeoGeo so much easier having a digital schematic displayed on a UHD LCD Monitor.

Thanks, Arno
 

Attachments

  • P1000919a.JPG
    P1000919a.JPG
    134.6 KB · Views: 10

maki

Edo Express Delivery Guy
Joined
Jan 1, 2022
Posts
332
I'm not a specialist

when you swapped the NEO-BUF, did the address change?
happened to me, swapped a broken one with another broke one, but the error address changed

you checked all the connection between the two Palette RAM ICs?
 

nvana

n00b
Joined
Jan 20, 2022
Posts
3
Afaik the address stayed the same after the Neo Buf swap.

The address lines of the Color RAM ICs (PA0..PA11) seem all connected correctly between the 2 chips.

I presume a logic analyzer would be the only way to debug these RAMs ?
 

maki

Edo Express Delivery Guy
Joined
Jan 1, 2022
Posts
332
when I swapped the NEO-BUF, only the address changed, I didn't notice until I checked a video or pic I luckily made from the original error message

these are basically multiplexers, hard to check IMO with logic analysers

I have a logic probe (cheap), if you look for a signal but get the 0V/low sound without "chatter", check with a MM if the voltage is above zero

my EEPROM burner (TL866II+) has the ability to do very basic tests on common logic and SRAM ICs, like "do the address lines work?", it did find two broken RAM, its not good enough to tell you if the RAM IC works at higher frequencies and the like, but if its really broken, it will find it

there is SRAM testers for retro stuff, check google

from the error it seems that bit 0 can't be properly read of the highest byte, so the upper palette RAM IC is affected, or probably the trace to it

so the connections between NEO-BUF and the CPU are also good?
 

nvana

n00b
Joined
Jan 20, 2022
Posts
3
when I swapped the NEO-BUF, only the address changed, I didn't notice until I checked a video or pic I luckily made from the original error message

these are basically multiplexers, hard to check IMO with logic analysers

I have a logic probe (cheap), if you look for a signal but get the 0V/low sound without "chatter", check with a MM if the voltage is above zero

my EEPROM burner (TL866II+) has the ability to do very basic tests on common logic and SRAM ICs, like "do the address lines work?", it did find two broken RAM, its not good enough to tell you if the RAM IC works at higher frequencies and the like, but if its really broken, it will find it

there is SRAM testers for retro stuff, check google

from the error it seems that bit 0 can't be properly read of the highest byte, so the upper palette RAM IC is affected, or probably the trace to it

so the connections between NEO-BUF and the CPU are also good?

Thanx! Could it be the highest bit of highest byte, as it shows write 5555, read 4555 ?

I wasn't able to check the connections between the CPU and the NeoBuf as the schematics I have of NeoGeo are too fuzzy for me to read when I print them...
 

maki

Edo Express Delivery Guy
Joined
Jan 1, 2022
Posts
332
lowest bit of the highest byte, as its the only one that can make odd numbers ;)

the NEO-BUF is a PITA to test, I just swap them, got a few rotten MVS PCBs for cheap
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
lowest bit of the highest byte, as its the only one that can make odd numbers ;)

the NEO-BUF is a PITA to test, I just swap them, got a few rotten MVS PCBs for cheap
I think Furrtek makes drop in replacements of the NEO-BUF.
 
Top