Topic: Neo Geo 4 Slot Repair Log  (Read 6957 times)

Author Message

0 Members and 1 Guest are viewing this topic.

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Neo Geo 4 Slot Repair Log
« on: September 02, 2010, 11:37:23 AM »
I don't get as much time these days for board repair as I used to but I did offer to take a look at Wallaroo1979's 4 slot Neo Geo which was giving a video RAM error at address 8000. Have seen this specific fault a few times before and address 8000 should relate to a couple of Sony CXK5316 SRAMs, which are a common problem chip for some reason.

I figured this would be a nice quick fix - I was wrong!

Unfortunately having given advice on the likely candidates on the forum the board ended up going to someone he knows in the electronics field, who removed the two SRAMs, as well as a mask ROM nearby that had a couple of spots of corrosion on a couple of pins (not a good enough reason to remove it if you ask me). Doubly unfortunately these chips had been taken off the board with a solder sucker (ie not a desoldering station) and the chips had taken a pounding. The mask ROM legs were covered in charred flux, bent to buggery and the 5V pin had been snapped clean off.



The PCB had some signs of a lengthy desolder, some plate thru holes looked a bit chewed up...



...and there were some dints in the board on the solderside. A quick sanity check of the ROM and RAM tracks showed that only the SRAM chips shared an address bus and that there was a hole in it. All address pins should be common between the 2 chips but A6 was not connected due to damage sustained during the desoldering, a wire link fixed this up.

Finally the board had also been liberally sprayed with what looked like WD40...



...I wasn't far wrong. From what wallaroo1979 told me (he was given the rest of the can by his contact) it's a water repellent and lubricant for switch gear, relays etc etc. The manufacturer makes no claims about its suitability for hosing down PCBs, water repellent is one thing but quite why a PCB with no moving parts would need a lubricant still escapes me; it makes the board look nice and clean but somewhat slippery.  :blink:

My main concern was the mask ROM, these are chips that are built with the data built in the chip fabric rather than as chips that need programming after manufacture. As such they can do away with some of the pins that an EPROM would need to support the programming function so they are often a non standard pinout. This mask was a T531000 chip which is a 1 meg mask (128x8) and was not supported by my eprom reader.

The two SRAMs were indeed totally shot, as I could not be sure the mask ROM was not also wrecked I fitted machined pin sockets for all three chips as the board would not survive any more desoldering even if it was done gently. I repaired the damaged5V  pin and reinstalled the mask ROM plus 2 TMM2018-35 SRAM chips borrowed from a board in my collection.



Powering it up I got this - a screen full of crap...



... if I pulled out one of the SRAMs to simulate the original fault I got a corrupt version of the Video RAM fault overlaid on the screen full of crap.



Wallaroo1979 confirmed that the original error message was just white text on a black screen which was what it should be, this board had acquired a new fault :(

When a NeoGeo board has no game cart it should boot to a cross hatch screen, in fact you can run a multi-slot board without the ROM board attached and it will act as if it has no cart even though an entire board is missing. The board would go through the standard bootup green screen process before getting to the above image, and it did look a bit like a mangled cross hatch, the characters in the error message also supported the idea that the board was booting but that the character data was bad, assuming the cross hatch was made up of individual characters which seemed feasible.

What was worrying was the fact that pulling the mask ROM out had no impact on this crap, which would mean either is isn't involved at all, or its absence causes this crap, which as it was present would indicate it was blown.

Poking a scope at the data, address and control lines on the mask ROM and the SRAMs didn't show anything obvious and as I was fairly sure the mask was dead I focussed on that. With a bit of googling I found that the pinout for the TC531000 was fairly similar to that of a 27c1001 EPROM. The mask ROM had fewer pins so if I moved the 5V input out beyond the end of the chip and moved address line 16 I could read the chip as if it was an EPROM. So by bending two legs and fitting some wires...



... I had a Frankenstein chip I could read. The subsequent dump came back as a perfect read of a healthy NeoGeo BIOS chip, not what I was expecting to be honest but it meant I didn't have to rig up a way to replace the mask rom with an EPROM.

While the chip was off the board I pulled out the two SRAMs and went round the sockets with the scope, everything looked normal except A3 (pin 7) on the mask. Buzzing through the lines I could prove that all the address lines on the mask ROM went to the SNK LSPC custom chip except A3, the address lines also go elsewhere but without the SRAMs and the mask on board the "elsewhere" was quiet so the scope showed up the lack of activity from the SNK custom on this one line.

After fitting another wire link to join this broken track to its via the board booted to the normal cross hatch.



:)

I was expecting that to be the end of it but after fitting the ROM board and whacking in a cartridge I got this...



:(

...a running game with all background and foreground graphics missing, just coloured blocks moving around and the text overlay. This is a classic sign of massive lumps of ROM data being missing, virtually all of it. Most likely candidates for losing everything would be either a bad ROM (proven not the case as the cart works in my 1 slot board), a blown buffer chip (i.e. all outputs shot), a missing enable signal somewhere or crappy cart connections losing the enable signal. After proving it wasn't a case of a bad cart connection the big question was where to start, normally you would start at the ROMs but you can't get to the ROMs easily as they are in the cart, even if you take the cart apart you find the sound PCB is in the way, all you cant get to is the solderside of the main rom cart PCB which isn't that great. If you remove the sound PCB the board wont boot so you cant get the fault to work on.

The only way forward was to blindly probe around the TTL on the ROM board. There are literally dozens of 74LS244s and 245s that string the 4 cart slots together and allow the isolation of one, working on the assumption that each slot would have the same set of chips and the fact the chippery was 16 rows deep  I assumed that each slot would have 4 rows of chips and the board did have four repeating patterns of chips, as I only had 1 cart in the board I only had to prove 1/4 of the chips. It didn't take that long to find something smelly either, the 2Y1 output on pin 9 of the 74LS244 at A3 was rather quiet...



...all others were busy and as the enable is common looked wrong, the input for this gate was on pin 11 and that was floating.

Briefly injecting noise into pin 11 gave me this...



...pretty much everything brought back but somewhat screwed up, which proved I was on the right trail. Pin 11 connected to all the other slot equivalent LS244s and ended up on A19 on the top left inter-board connector, I traced this across the solderside of the entire board to a tiny via between the legs of the 74LS244 at R11 on the CPU board. From there it seemed to go nowhere, i scraped some lacquer off a couple of likely tracks coming out from the chip but they were not it so I was faced with removing the chip to see where the track went or chancing my luck with some logic. Seeing as this was one line of 8 on the 74LS244, if all other lines went to pins on the 74LS244 at the far end except one, and one output of that chip didnt connect to the edge connector it was likely that that was the missing track. Jackpot - all outputs of the 74LS244 went to the edge connector except pin 5, which seemed to go nowhere.

I tried to solder a link wire to the via but I could not get a connection, am pretty sure the via had corroded away to nothing so I scraped some lacquer off the track upstream and fitted a link between pin 5 and that track. A quick sanity check with the continuity meter that everything was connected and not shorted out (coz them tracks are damn thin to be soldering to) and...



...Fixed!

Have confirmed all 4 slots work and the board cycles through the attract mode of all installed carts! Just needs some lacquering and she ready for a road trip back to wallaroo1979.

Its likely that other vias in the area are in a bad way too but the only way to tell is to disturb them which is likely to cause a fault even if one wasn't going to occur for years, or ever. I should really record connectivity pattern for the other lines to that chip in case this board starts giving graphics issues in the future.
« Last Edit: September 03, 2010, 01:28:24 PM by Womble »
Sic Transit Gloria Atari

Offline AndyRCM

  • >=))))º> GO FEED THE FISH! <º((((=<
  • Administrator
  • Amiga 4000
  • ******
  • Posts: 9651
  • Kudos 50
  • Gender: Male
  • Manic Jet Set Willy
    • View Profile
    • Retro Computer Museum
Re: Neo Geo 4 Slot Repair Log
« Reply #1 on: September 02, 2010, 12:11:10 PM »
Bloody hell mate - really nice work indeed. That was a hard one! :)

A
"I could see the faces of those who led pissing themselves laughing" - Funeral Pyre by The Jam

Offline porchy

  • Moderator
  • Amiga 4000
  • ******
  • Posts: 825
  • Kudos 33
  • Gender: Male
    • View Profile
    • JAMMArcade
Re: Neo Geo 4 Slot Repair Log
« Reply #2 on: September 02, 2010, 12:52:20 PM »
good stuff.
ive got a couple sets of spare maskroms off neo geo boards should you ever need them in the future

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Re: Neo Geo 4 Slot Repair Log
« Reply #3 on: September 03, 2010, 12:18:52 AM »
Bloody hell mate - really nice work indeed. That was a hard one! :)

It was a bit of a pain, the guy was asking for help on the AA forum for a while but I was a bit reticent to get involved, still got board from other folk awaiting my attention. I was pretty annoyed that between me posting the likely candidate chips on the thread and offering to help it had gone to someone who he stated didn't think could fix it. Afterall you don't send your car to someone who says they wont be able to fix it but that they could take it apart for you if you like. Half the faults were from the other guys work, and that oil all over the board... had to keep washing my hands. Still it makes for some nice glossy photos I guess :)

ive got a couple sets of spare maskroms off neo geo boards should you ever need them in the future

Cool, thanks will bear that in mind  :)

Actually the "don't get the time anymore" is not strictly true, I just don't make the time anymore. For some reason it doesn't appeal as much as it used to, burnout perhaps, I did get buried in boards from folk. My interest has pretty majorly swung away from Arcadia back to retro computers which is surprising. Anyone know if its possible to buy a spare tape deck cover for a CPC464? ;)
« Last Edit: September 03, 2010, 01:19:36 AM by Womble »
Sic Transit Gloria Atari

Offline AndyRCM

  • >=))))º> GO FEED THE FISH! <º((((=<
  • Administrator
  • Amiga 4000
  • ******
  • Posts: 9651
  • Kudos 50
  • Gender: Male
  • Manic Jet Set Willy
    • View Profile
    • Retro Computer Museum
Re: Neo Geo 4 Slot Repair Log
« Reply #4 on: September 03, 2010, 07:51:23 AM »
Actually the "don't get the time anymore" is not strictly true, I just don't make the time anymore. For some reason it doesn't appeal as much as it used to, burnout perhaps, I did get buried in boards from folk. My interest has pretty majorly swung away from Arcadia back to retro computers which is surprising. Anyone know if its possible to buy a spare tape deck cover for a CPC464? ;)

I would like to think that RCM has had something to do with that! ;)

A
"I could see the faces of those who led pissing themselves laughing" - Funeral Pyre by The Jam

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Re: Neo Geo 4 Slot Repair Log
« Reply #5 on: September 03, 2010, 10:35:12 AM »
I think visiting the RCM groto broke me, haven't really had the arcade urge since I got back from blighty.
« Last Edit: September 03, 2010, 01:26:28 PM by Womble »
Sic Transit Gloria Atari

Offline porchy

  • Moderator
  • Amiga 4000
  • ******
  • Posts: 825
  • Kudos 33
  • Gender: Male
    • View Profile
    • JAMMArcade
Re: Neo Geo 4 Slot Repair Log
« Reply #6 on: September 03, 2010, 10:37:35 AM »
I think visiting the RCM Groton broke me, haven't really had the arcade urge since I got back from blighty.

in that case, fancy selling me your Rainbow Islands and New Zealand Story boards really cheap??  :D

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Re: Neo Geo 4 Slot Repair Log
« Reply #7 on: September 04, 2010, 01:00:57 AM »
Gasp, my preciouseseseseses????
Sic Transit Gloria Atari

Offline porchy

  • Moderator
  • Amiga 4000
  • ******
  • Posts: 825
  • Kudos 33
  • Gender: Male
    • View Profile
    • JAMMArcade
Re: Neo Geo 4 Slot Repair Log
« Reply #8 on: September 04, 2010, 09:42:03 AM »
Gasp, my preciouseseseseses????

ill take that as a yes, ill open the bidding at 10 english pounds for the pair?

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Re: Neo Geo 4 Slot Repair Log
« Reply #9 on: September 05, 2010, 02:50:11 AM »
Gasp, my preciouseseseseses????

ill take that as a yes, ill open the bidding at 10 english pounds for the pair?

But but but I have no use for English pounds!!
Sic Transit Gloria Atari

Offline porchy

  • Moderator
  • Amiga 4000
  • ******
  • Posts: 825
  • Kudos 33
  • Gender: Male
    • View Profile
    • JAMMArcade
Re: Neo Geo 4 Slot Repair Log
« Reply #10 on: September 05, 2010, 11:43:45 AM »
then how about a boxset of Tintin cartoons and a chewed biro?

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Re: Neo Geo 4 Slot Repair Log
« Reply #11 on: September 05, 2010, 01:55:52 PM »
What colour biro?
Sic Transit Gloria Atari

Offline porchy

  • Moderator
  • Amiga 4000
  • ******
  • Posts: 825
  • Kudos 33
  • Gender: Male
    • View Profile
    • JAMMArcade
Re: Neo Geo 4 Slot Repair Log
« Reply #12 on: September 05, 2010, 01:57:55 PM »
thats where you have a choice, black or blue, there is a red one but its not pre chewed, you would have to chew it yourself

Offline Womble

  • Amiga 4000
  • ******
  • Posts: 1190
  • Kudos 33
  • Gender: Male
    • View Profile
    • Aussie Arcade
Re: Neo Geo 4 Slot Repair Log
« Reply #13 on: September 06, 2010, 03:56:33 AM »
thats where you have a choice, black or blue, there is a red one but its not pre chewed, you would have to chew it yourself

Chew it myself???  :o

Sounds dangerous, I don't like it.
Sic Transit Gloria Atari