Neo Geo MVS 1FZ BRAM Dead Output (Upper) Address D000000

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
Hi all!
Learning as I go and loving it!

TL;DR Neo Geo MVS 1FZ is driving me mad! I have replaced 5 ICs and its still not working... Where to next? (Logic probe is in the mail)

1) Got it as a parts console - first booted up with Lower BRAM Error - wouldn't let me get to the other menu via A+B+C+D+ST+SEL. (It initially said: Backup RAM error ADdress 00D000000 Write 5555 Read 0800 but that has since been resolved and now I am stuck with a new error)
2) Replaced the RAM for Upper and Lower RAM with new 62256s (Tried doing it with a donor board first and same issue so I replaced with new)
3) Cool now I was able to boot into the Diagnostic bios menu and see other tests - Oh and lower BRAM error went away but now I have an Upper RAM issue
4) Read a bunch up on: http://smkdan.eludevisibility.org/neo/diag/#causes. (Traced it out and I have perfect connectivity to the C1)
5) Read a bunch on the Wiki page: https://wiki.neogeodev.org/index.php?title=Backup_RAM
6) Watched one of the best Youtubers for Neo Geo anything @GadgetUK+++++ on a similar issue:
7) Replaced U17 from reading this about the 74HC32A https://www.neo-geo.com/forums/index.php?threads/mv-1fz-u17-pinout.235924/ as this was what was needed to fix my Backup RAM error. Still had a Backup RAM error so I replaced IC1 as well from the post above
8) Replaced IC1 as well because for good measure
9) Then read up on a transistor issue on Q1 - replaced that too as I was reading nothing on the transistor when on and or off.
10) Replaced the U7, as well as the schematics, looked as it was part of the flow of the BRAM
11) Looked everywhere for broken traces and measured continuity and I am getting proper continuity

11) Ughghhhhghghghgh I think the only thing that has not been replaced on the BRAM circuit is the C1 at this point but I have connectivity so I am not sure.

Anything else I should try before this thing goes in the bucket of shame till I need my RAM ICs back? haha :P
 

Attachments

  • a2.jpeg
    a2.jpeg
    313.8 KB · Views: 12
  • a1.jpeg
    a1.jpeg
    939.1 KB · Views: 12
  • a3.jpeg
    a3.jpeg
    599.2 KB · Views: 14
  • a4.jpeg
    a4.jpeg
    408.8 KB · Views: 14
  • a5.jpeg
    a5.jpeg
    1.2 MB · Views: 15
Last edited:

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
The BRAM DEAD OUTPUT (UPPER) is the error you should focus on. This is indicating the chip isnt getting enabled when the CPU is trying to do a read. Check on the OE and CE lines of the backup ram chips.

Also fyi, automatic testing at startup does the following for ram:

- output enable test
- writable test
- data test
- address test

When you start a ram test from the menu system its only looping over the following 2 tests

- data test
- address test

The underlying problem is the dead output, which the manual test is reporting as a data error since it doesn't do the output enable test.
 

maki

Edo Express Delivery Guy
Joined
Jan 1, 2022
Posts
332
here is he pinout of the backup ram for the MV1FZ, should be very close to yours IMO

I've replaced many ICs myself and then found out its a trace :)
 

Attachments

  • MV1FZ Backup RAM.jpg
    MV1FZ Backup RAM.jpg
    1.5 MB · Views: 31

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
I am going to take them off and see if its a trace issue as all the continuity checks passed from a quick pass I did a week back.
Going to need to do it again then see if I need to remove the chips and inspect for damage harder underneath?
 

maki

Edo Express Delivery Guy
Joined
Jan 1, 2022
Posts
332
Going to need to do it again then see if I need to remove the chips and inspect for damage harder underneath?
If you're comfortable taking them off and on again, its a good opportunity to swap them and see if the error changes (broken traces vs IC), just make sure to mark them so you know which one was which.
Did you check of OE and CE signals are present and as excepted?
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
I did check the OE and CE signals and got what was expected so I was a bit confused as to what to do next.

Think I will take them off and inspect
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
Reflowed the back up rams and installed a new 74hc32A and I am no longer getting the bram error :)

Progress over perfection ❤️

Now I am getting a new error
Thinking the wrams near the brams?

TheY were replaced with used ram chips so it might be that?

I am going to need to burn the new diag bios so it can tell me if which ram chip but I am curious what that error means?
I read something on hex and translating it but will read up so I can learn.

Any pointers before I read more on this?
Ps thanks so much! Hope this is fun for you as it is for me. Love learning
 

Attachments

  • 7775E1F8-819F-4A4F-821D-A0672C3AD9D7.jpeg
    7775E1F8-819F-4A4F-821D-A0672C3AD9D7.jpeg
    948.2 KB · Views: 14
  • E65FFC88-FC55-4ED1-809E-7FC65CFB7ED5.jpeg
    E65FFC88-FC55-4ED1-809E-7FC65CFB7ED5.jpeg
    2.5 MB · Views: 14

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
Its actually a step backwards unfortunately. The wram test happens before the bram test.

Easiest way is using the built in calculator for whatever OS you run, it should have a programmer mode . This for example is the top part of the calc from OSX when in programming mode.

0x5555.png0x1555.png

There is only a 1 bit difference between them.
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
Oh dang lol.
Well, hmm no harm here, just means I learned that wram comes before bram test so I can fix it.


Yes! Found the link

So it’s something on the first data line causing me grief on a WRAM and it being the first digit 2 digits of the expected (Aka AA of the AA18 example in the link above) or in my case 15 of the 1555 it means my upper vram is causing an error.

Then it would mean I need to inspect:
Which tracks as that’s the one I should have pulled off and redone.
 

Attachments

  • FFDE5A1F-E29D-41C7-AB82-B8251BA9AD87.png
    FFDE5A1F-E29D-41C7-AB82-B8251BA9AD87.png
    380.9 KB · Views: 10
  • 5E536E95-2FB0-4E20-879F-6DE39BA8A4E9.png
    5E536E95-2FB0-4E20-879F-6DE39BA8A4E9.png
    380.7 KB · Views: 10
  • 660268E4-F4CE-429A-AB84-D71A8998BFF0.jpeg
    660268E4-F4CE-429A-AB84-D71A8998BFF0.jpeg
    443.8 KB · Views: 16
Last edited:

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
Well, hmm no harm here, just means I learned that wram comes before bram test so I can fix it.

This page gives you the list of the tests and the order

https://github.com/jwestfall69/neogeo-diag-bios/blob/master/docs/tests.md

You can also identify the specific data line causing the problem.

Code:
0x5555         0101 0101 0101 0101
Data Lines     7654 3210 7654 3210
              |  Upper  |  Lower  |
              
0x1555         0001 0101 0101 0101
Data Lines     7654 3210 7654 3210
              |  Upper  |  Lower  |

So its data line 6 of the upper chip that is problematic. Note that the upper chip's data lines D0-D7 connect to the cpu's D8-D15 data lines. The issue could be a trace issue or maybe the ram chip is bad.
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
Its alive! well sort of haha Need some help again <3

Sorry been a bit since I replied to this thread. Fixed like 4 other Neo Geo MVS consoles and put this one in the bin of shame while I worked up the patience to attempt it again.Bad traces on the upper RAM lines.

Took off the RAM and noticed that the pads were basically eaten away.
I mean, I did replace and replace like 4 times.. still no excuse as I got messy one removal. (prob was loose a while before I touched it too? IDK)


TL;DR I get a 100% pass on things with no game in but Z80 testing gives me the following Blue screen of death. Next steps?

1) I have another FZS I have replaced a bunch of parts on and even replaced the Z80 and got it to stop saying its Dead, but I have a similar Blue screen issue.
2) No threads out there on this Blue screen of death, unless I am not seeing any
3) Any ideas or suggestions? Z80 has been cleaned with a fiberglass pen and no trace damage from what I saw in a quick Multi-meter probe

@ack or @maki

PS thank you two so much! You really have helped me level up my soldering skills and hardware skills over all.
I may now have the confidence to attempt to repair the 6 AES consoles I have sitting in the bin of Shame. lol (I have fixed 2 AES bad Unibios prob cases before but the AES scares me because AE$=Pressure)
 

Attachments

  • image3.jpeg
    image3.jpeg
    119.3 KB · Views: 9
  • image1.jpeg
    image1.jpeg
    64.7 KB · Views: 9
  • image2.jpeg
    image2.jpeg
    110.7 KB · Views: 9
  • image0.jpeg
    image0.jpeg
    72.2 KB · Views: 10

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
The blue screen isn't related to the z80, which only does sound stuff. When the diag bios does the slot switch to make the diag M1 rom active, the switch will also cause the board to switch from the onboard SFIX rom to the S1 on the cart. The SFIX/S1 roms contain the chars that get used to draw the text on screen. If the issue is happening on multiple boards, its an issue with your diag cart's S1 rom or the address/data lines to it.
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
Ahh, this is my garbage cart aka I wouldn’t be upset if it got damaged. That makes sense that I’d see similar behavior as I always use this cart to test things. (It’s super bootleg and wouldn’t recommend it to anyone haha)

I am going to test with a real game soon.

Thanks for letting me know, I did some reading up on the diag bios page and it seems I need to have a CHA board, and having read up, seems a few games have this pcb type: http://www.arcade-collector.com/search-neogeo-board-cha-8m

Yay! I have super spy which may be a good test cart for this.

Hopefully that’s what’s up.
Oh I also got an M1 bios chip to test things with as well, probably going to pop that in and use a legit CHA game and report back :)

Thanks again! Let me know if I am off base haha
 

maki

Edo Express Delivery Guy
Joined
Jan 1, 2022
Posts
332
Took off the RAM and noticed that the pads were basically eaten away.
I mean, I did replace and replace like 4 times.. still no excuse as I got messy one removal. (prob was loose a while before I touched it too? IDK)
I can't help with the blue screen of death problem, but this one I know, happened to me too :)

The pads are being held on the PCB with epoxy glue, on Neo Geos I made the experience that the pads will come off easily if "too much heat" is applied, especially when mechanical stress is applied (desoldering gun recoiling and hitting the traces).
The more you solder on these pads, the more likely they are to lift. These PCBs are very fragile, some traces will come of if you stare to long at them ;)

I'm making sure my soldering temps are between 300C and 320C for these PCBs, still fragile though, then there is "low melt solder" with Bismuth, this can take down the de-soldering temp to below 100C, that can be quite safe for the PCB.
However, one must (!) remove all traces of the low melt solder for the PCB and the soldering tip, lots of leaded solder works fine for me, then remove it with desolder-braid, few passes and you're good, just be careful when removing the "contaminated" solder, don't rub too hard on the desolder braid.

Sure one can get the totally overpriced ChipQuik as low melt bismuth solder, but pure Bismuth is sold on ebay for a fraction of the price and just needs some (leaded or unleaded) solder added to it (about half or more), this is were left over solder from the desolder pump/gun comes in handy.

I bought this: https://www.ebay.com/itm/164031457104
however, the seller and me are both in Germany, you might want to look for a seller in your country.
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
Tried the z80 test with a different cart. Yay! Z80 testing is now working. Idky but I just get scared using og carts when it comes to testing, I’d rather buy another AliExpress cart any day than damage an original.

I have 2 fz consoles that I repaired up to this point and they get the same message in the pic below.

I did replace the z80 in one of them and no change.
Do you think the z80 ram could be causing issues?

I tried to see if I could source replacements but I didn’t see that exact model I was looking for.

Plus it’s skinny 24pin which I saw a lot of discontinued messages in digikey’s inventory

Let me know your thoughts as I can get out and prob things but I almost want to socket the ram and see if that’s being a pain
 

Attachments

  • BAAE9082-2A1F-43E5-84BB-C5ABBE518896.jpeg
    BAAE9082-2A1F-43E5-84BB-C5ABBE518896.jpeg
    1.7 MB · Views: 9
  • 66D6DAEC-12AA-4302-AC4A-008C860BBB47.jpeg
    66D6DAEC-12AA-4302-AC4A-008C860BBB47.jpeg
    2.5 MB · Views: 10
Last edited:

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
First thing I would do is verify you have a fully working diag cart by trying it on a known good board.
 

Tengugurl

Cheng's Errand Boy
Joined
Feb 9, 2017
Posts
111
@ack I am so new to this lol! I thought I just use a regular cart and I was so mistaken haha

I found an awesome member here to sell me a diag cart and tada! It’s fixed.
Omg you all have been so helpful ++++++++<3

For anyone reading this in the future lessons learned ahead:

1) don’t hear the pads up too hot. Else you’ll be doing trace work
2) don’t replace things that don’t need it. Be methodical
3) don’t be cheap. If you don’t have the right parts. Don’t guess on sram and risk damaging traces
4) patience patience patience
5) the community on here are amazing and truly a bunch of neo geo geniuses
6) you can fix things if you have the right tools Aka a diag cart not a standard cart


Thank you all!
6/10 neo geo mvs pcbs fixed.
Going to keep on practicing till I can be the one to reply with helpful tricks and tips

6617E852-4E52-4749-892D-091B27E5B1BD.jpeg
 
Top