MV-1FZ weird problem (BIOS and BACKUP RAM ERROR)

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Hey Guys.

So I have another MV-1FZ that makes me scratch my head. It seems to have Backup RAM Problems but as well will report a Bios Error with no real Address when trying to boot it up with a UniBios.

So I checked continuity from all A and D Lines from the Work and Backup RAM to the 68k.

Also I checked continuity from all A and D Lines from the System ROM to the 68k.

Those measurements where all fine.

The SNK Diagnostics Bios will give a Bios Checksum Error as well.

And as you can also see you will see those weird patterns on the screen.

Judging by how I received the board someone might have pulled it out of the pacific ocean or something as it has residue of dried up dirt/water (kind of).

I did a dish washing and cleaned the whole board with IPA and a brush and reflowed all of the parts with the same results.

Also the Backup RAM have been replaced as well as the HC32. There is no battery damage.

Maybe someone can help me out a little bit before I start replacing part after part and taking off the LSPC2-(A3) and the B1.

Thanks and have a good day
 

Attachments

  • 1fz_1.JPG
    1fz_1.JPG
    182.4 KB · Views: 74
  • 1fz_2.JPG
    1fz_2.JPG
    213.7 KB · Views: 76
  • 1fz_3.JPG
    1fz_3.JPG
    112.5 KB · Views: 75
  • 1fz_4.JPG
    1fz_4.JPG
    108.8 KB · Views: 76

Razoola

Divine Hand of the UniBIOS,
Staff member
20 Year Member
Joined
Nov 12, 2002
Posts
4,662
The bios error simply means the bios is failing its internal checksum. Providing your sure the EPROM is programmed correctly the error is likely to be on one of the higher address lines in relation. Can you get to the unibios main menu or does it crash before then?
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
The bios error simply means the bios is failing its internal checksum. Providing your sure the EPROM is programmed correctly the error is likely to be on one of the higher address lines in relation. Can you get to the unibios main menu or does it crash before then?

If I hold own ABC I come to the main menu and can select the menu items to change the Region and Mode.

I cant go to the ingame menu as the game will not start.

I also noticed that some of the Graphic Glitches go away whenever I hold down a button on the Controller.
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Its really strange - could it be that the CPU is broken ?

Any hints what to test next - this board really seems to have a lot of trouble

The BIOS works on my MV-1C so the EPROM is ok
 

Attachments

  • bios_.JPG
    bios_.JPG
    118 KB · Views: 62
Last edited:

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
So I replaced some more things now

* Palette RAM
* Work RAM
* Backup RAM

When I plugged the Diagnostic Bios in I also get a Bios CRC Error.
When running the test manually I will pass the VRAM and Calendar Tests - however the BRAM and Palette Test will bring the following errors (see pics).

Also the Color Bars look messed up
 

Attachments

  • bram1.JPG
    bram1.JPG
    243.4 KB · Views: 67
  • colorbars.JPG
    colorbars.JPG
    144.8 KB · Views: 64
  • palette.JPG
    palette.JPG
    154 KB · Views: 63

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Are the glitchy pixels fixed or moving all the time ? Looks like writes to palette RAM are happening when they shouldn't.
Is NEO-C1 getting A17~A21, A22I and A23I ? Guessing A22I and A23I are ok since the BIOS starts, so suspecting the lower lines.

The color bars look ok expect for colors 3, 7, 11, 15 which seem to have been overwritten with garbage.

See if NEO-C1 pin 16 (PAL signal) is stuck low or going crazy while on steady screens. That would explain palette RAM corruption and checksum errors because of palette RAM data being placed on the bus when it shouldn't.
 
Last edited:

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Are the glitchy pixels fixed or moving all the time ? Looks like writes to palette RAM are happening when they shouldn't.
Is NEO-C1 getting A17~A21, A22I and A23I ? Guessing A22I and A23I are ok since the BIOS starts, so suspecting the lower lines.

The color bars look ok expect for colors 3, 7, 11, 15 which seem to have been overwritten with garbage.

See if NEO-C1 pin 16 (PAL signal) is stuck low or going crazy while on steady screens. That would explain palette RAM corruption and checksum errors because of palette RAM data being placed on the bus when it shouldn't.

The glitchy pixels are moving all the time!

I got A17-A21 from the 68k to the NEO-C1
A22I and A23I is coming from the NEO-E0

NEO-C1 Pin 16 is stuck HIGH.

Any more hints ? Thanks
 

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Check that NEO-C1 pin 16 goes to the LS32 "U1" pin 12 next to RAM10, and also pin 19 of both 74LS245 "U8" and "U9" next to the JAMMA edge.
Also check that 68k pin 9 (RW) goes to that same LS32 pin 13.

If everything is connected ok. Replacing that LS32 and/or both LS245 might help.
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
So I got NEO-C1 Pin 16 to LS32 pin 12 - as well as LS245 pin 19 (both IC's)
68k Pin 9 to LS32 pin 13 is also ok.

Will replace those three IC's now - fingers crossed
 

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Fingers crossed here too. I'll try to guess exactly what's going on from your pictures and the diag bios code.
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Fingers crossed here too. I'll try to guess exactly what's going on from your pictures and the diag bios code.

Replaced those three IC's however exactly the same problem :(

I think it's strange that the glitches get less when I hold down a button.. this might be related to NEO-C1.
What is the chance that NEO-C1 is faulty ?
 
Last edited:

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Hmm :/
Already had a faulty NEO-C1 on a MV1FZ which was causing Z80 errors, so apparently it can happen.

The glitches might change when you hold a button because it makes the code take a different path.

Is 68k pin 9 going to pin 1 of both LS245 also ?
 
Last edited:

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Hmm :/
Already had a faulty NEO-C1 on a MV1FZ which was causing Z80 errors, so apparently it can happen.

The glitches might change when you hold a button because it makes the code take a different path.

Is 68k pin 9 going to pin 1 of both LS245 also ?


Yes I have continuity from 68k Pin 9 to LS245 Pin 1 (both ICs)
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Replaced the NEO-C1 with exactly the same results like before.. So its probably not the NEO-C1 :(
hmm could it be the NEO-D0 ? It's also handling Joypad and 68k Addresses ?
 

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
NEO-D0 only generates some clocks, gives the outputs for the controllers and deals with the Z80.
Can't see how it would be related to the problems your board has but since I could have sworn it was the palette RAM buffers going nuts, I really don't know in which direction to point now :/
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
NEO-D0 only generates some clocks, gives the outputs for the controllers and deals with the Z80.
Can't see how it would be related to the problems your board has but since I could have sworn it was the palette RAM buffers going nuts, I really don't know in which direction to point now :/

So just to get some more pointers - the Backup RAM will not work even though there is continuity and I replaced them with other chips several times.
I may replace the 68k but it seems that at this point I may even face multiple part damage that will not make troubleshooting much easier..
 

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Just got another MV1FZ in which had similar scrolling glitches and was stuck on a backup RAM error at $D0000C.
The 74HC259 "U7" had toggling outputs when it wasn't addressed, replaced it and the slot is now working again.
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Just got another MV1FZ in which had similar scrolling glitches and was stuck on a backup RAM error at $D0000C.
The 74HC259 "U7" had toggling outputs when it wasn't addressed, replaced it and the slot is now working again.

Thanks once again man !!
The U7 was a very good tip it fixed my Graphical problems. Then I had some watchdog errors left and replaced U1 and U6 again.
Also I found a broken trace on NEO-D0 (SDMWR Pin).

Then the Time in the UNIBIOS was really increasing fast (1 hour / second almost) so I replaced the 4990AG and the Crystal.

That fixed the Time going up fast however the date and time will be locked to 00/00/00 now and not change.
Also I got no sound ?
What could cause those two problems ?
 

Attachments

  • D0.JPG
    D0.JPG
    253.7 KB · Views: 51
  • diagnos2.jpg
    diagnos2.jpg
    90.2 KB · Views: 52
  • test_snk.JPG
    test_snk.JPG
    211.7 KB · Views: 53

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Glad it helped :)
I have no clue why those 259 fail that way. The one on my board had its SHADOW output toggling all the time and caused the pixel snow.
Maybe yours was doing that on the palette bank output, causing something similar but more colorful :p

Are you sure the new RTC chip is ok ? I guess the oscillator is running, otherwise the diag bios would throw an error (I think ?).
Maybe a cut trace between the RTC and NEO-F0 ?

No sound could be caused by many things. Since you saw a cut trace around NEO-D0, maybe there are more ?
No Z80 error with the stock bios ? Maybe some data lines are cut between the Z80 and NEO-C1 (SDD* pins).
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Glad it helped :)
I have no clue why those 259 fail that way. The one on my board had its SHADOW output toggling all the time and caused the pixel snow.
Maybe yours was doing that on the palette bank output, causing something similar but more colorful :p

Are you sure the new RTC chip is ok ? I guess the oscillator is running, otherwise the diag bios would throw an error (I think ?).
Maybe a cut trace between the RTC and NEO-F0 ?

No sound could be caused by many things. Since you saw a cut trace around NEO-D0, maybe there are more ?
No Z80 error with the stock bios ? Maybe some data lines are cut between the Z80 and NEO-C1 (SDD* pins).

Now my Board is dead again.. Hmm.. the unibios will be stuck with a pattern. the diagnostic bios will show a VRAM Data Error 0000000 and when trying to hold ABCD the diagnostic bios will freeze on the main menu.

Throws me back a couple of steps. Replaced the VRAM but get the same error so I guess another broken trace or something.... would have been too easy..
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
So this Error should be the "normal" Video RAM - but it's strange as I replaced them already and I can't come this far with the unibios (unibios will lock with a garbled screen without watchdog - totally frozen).

The Diagnostics Bios will show the same VRAM Error and if you want to go to the Menu Holding down ABCD it will freeze in the Menu.

Some strange things going on here. I wonder why this suddenly happened as I did not really do much from the point where it will play games (without sound) to this VRAM and Freezing error now.

I checked continuity from the VRAM to the LSPC2 - all lines seem ok.

Is there any other IC that could be causing me this trouble ?
 

Attachments

  • ramerror2.JPG
    ramerror2.JPG
    46.9 KB · Views: 41

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Pretty much stuck with the Video RAM Error.

Just to clarify those should be the Lower VRAM (LSPC2-A2 / B0~B14: Low VRAM address bus / E0~E15: Low VRAM data bus) ?
Well continuity on all pins seems good and I replaced all the VRAMs (also Palette to be sure) without getting anything else than this error - it's really making me sick as the board was running before.

I reflowed the whole board again but I always end up with the same error (and still the SNK Diagnostic will freeze in the Main Menu).

Not sure how to continue on this board.. maybe I will have more luck tomorrow
 

furrtek

Super Sidekicks 2
Joined
Feb 6, 2011
Posts
106
Strange once again. The low VRAM is obviously good since the error text can be displayed.
Maybe the writes are working but not the reads. Is the LSPOE trace ok between NEO-C1 and LSPC ?
 

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
Strange once again. The low VRAM is obviously good since the error text can be displayed.
Maybe the writes are working but not the reads. Is the LSPOE trace ok between NEO-C1 and LSPC ?

Yes both traces are good LSPOE and LSPWE going to the NEO-C1.
I also think it's strange why it is freezing in the diagnostic..

Tried firing it up today - Now it does not give any video output and the LSPC2-A2 is getting warm..
Will put it inside the dishwasher and then look again after work (this board makes me feel like "meh"...)
 
Last edited:

RyuX

Crossed Swords Squire
Joined
Sep 19, 2018
Posts
179
So after some dishwasher and drying action I am back to my good old VRAM ERROR (ADDRESS 000000)

I replaced the NEO-ZMC2 (without any specific reason other than having a spare one) - completely the same.
So my last resort will be LSPC2-A2 and NEO-B1 replacement but I will postpone this decision as long as possible because usually this is some real work and you end up with more broken traces than before even if you are very careful with the reflow station.

Anyone having a pinout for all VRAM related connection between ICs that I can check ? I only made myself the LSPC2 and 68k cheatsheet but I don't have much on all the other stuff.. :(
 
Top