MV1-1 glitched Text - HELP!

F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Hello everyone,

I´ve got another MV1-1 board to fix.
First it was watchdogging and I used the logic probe to see pulsing on the pin 20. The Bios chip was dead. Using my other bios chips or Unibios/Dan diagnostics it booted to an error screen. The text was all weird but I could read: "Vram 2K dead output (lower)". I replaced both Fast Vrams because they were defective. The upper one showed "unwritable".

Then I got a Vram adress error, and replaced both Vram Chips. The board booted to: "All tests passed", But the text is all messed up, and I got no sound. When I boot a cartridge, the game graphics and in game text are perfect! I only dont have sound.

I ran the Z80 tests using a cartridge on the rising board, and it said something like "SM1 seems unresponsive".

Sometimes I see activity in the screen. Single pixels turning on or off randomly too… Weird. On my other boards this disappeared with the new VRAM.

Here are the pictures of both situations: With unibios crosshatch, and SNK Dan Diag.:

IMG_20190807_175831.jpg


IMG_20190807_175929.jpg

Does anyone have any Idea where and how to start troubleshooting this?


thank you all!


Regards
 
ack

ack

Mickey's Coach
15 Year Member
Joined
Apr 9, 2009
Posts
577
The 2 pictures give the impressions you have a stuck address line on the SFIX rom causing the wrong tile to load.

https://wiki.neogeodev.org/images/archive/1/15/20170219113500!Fixref01.png

For example in the unibios shot. "S" in the soft reset line should be the single tile "S" (0x0053), while what appears to be being displayed is the top part of the double tall "S" (0x0153). That would point to the 9th address line being stuck (pin 21? on the SFIX).
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Thank you for the help.

Indeed I found pin 20 Stuck Low. All the others are pulsing, except Vcc and GND. Is this the way to go? Try to understand why 20 is stuck?
 
ack

ack

Mickey's Coach
15 Year Member
Joined
Apr 9, 2009
Posts
577
The docs I have show pin 20 is Chip Enable, so I believe it will be low unless the cart's S1 chip is active.
 
ack

ack

Mickey's Coach
15 Year Member
Joined
Apr 9, 2009
Posts
577
I think what I said before is incorrect. The 0x0053 and 0x0153 refer to the tile number and not the memory address on the SFIX rom. It possible there could be something up with whatever is converting the tile number into requests to the SFIX rom. I think the LSPC-A0 handles that, but not 100% sure.

Re-reading your original post, with sprites and fixed working with a cart, it could just be you have a bad SFIX rom.
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Hi,

I swapped the SFIX rom chips with one from another board (unsure if it was good or not), but the result is exactly the same.
Maybe the LSPC-0 Is crazy? :conf:

thanks!
 
BIG BEAR

BIG BEAR

SHOCKbox Developer,
20 Year Member
Joined
Dec 14, 2001
Posts
8,383
check the LO chip. I had the same problem and ended up having to replace it if I recall correctly.
BB
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Hi,

I replaced the L0 Chip and the problem remains exactly the same.
Could any of the logic chips cause this problem?

thanks!
 
BIG BEAR

BIG BEAR

SHOCKbox Developer,
20 Year Member
Joined
Dec 14, 2001
Posts
8,383
Check the NEO-IO ...
Is this the NEO-MVH MV1 with the two ports for the cartridge interface,one on the left and one on the right or is it the MV1 with the three ports??
I ask because NEO-MVH MV1 has the wires attached the LS32 and the other one does not.
BB
 
Last edited:
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Hello Guys,

By flexing the board pressing the NEO-I0 chip, the text suddently appeared the right way! :D

But I reflowed the NEO-I0 to no result. I found some inputs floating between the LS244 at E1 and the LS32 at D1. At least now I know it´s due to a bad connection/trace somehow! i´ll try to find it and kill it !!

I´ll be back soon, with good news , I hope…

Thank you all ;)
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
:( couldn´t find anything. The only thing that brought the text back was twisting the SFIX. But there is a problem. I removed the SFIX and improved the connections on the traces. Now the problem persists and I Can´t make the text good again no matter how much I twist the board or the SFIX chip. I think by improving the traces, I made the problem worse.
 
BIG BEAR

BIG BEAR

SHOCKbox Developer,
20 Year Member
Joined
Dec 14, 2001
Posts
8,383
At least you know there is a bad trace/track somewhere and it's not a bad component... I don't have access to my boards yet so I cannot go into any detail atm...maybe in a couple of hours...
BB
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Thank you!

I´ll get some rest.

What makes me mad, is that I have good continuity between all the SFIX pins now. I wonder what was making the text appear good.
Tested with 2 SFIX chips, same result. could get good text bending the board/twisting the SFIX´s. Watching the board against a light shows nothing wrong. I was wondering if I had a short on the SFIX vias, and twisting it was taking it away…

Sometimes bending the board makes it reboot, so I must still have any inttermittent connection…
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Hello guys,

Now bending the board has no effect on anything.
Ive continuity tested a lot of things and I cannot get this board to boot to correct text.

Some tests I did gave me some results:
Shorting pin 26 and 27 on the SFIX made the text look right but with some artifacts.
Powering the board with pin 26 out of the socket made it boot with the right text size, but with some artifacts. After I approached my hand of the chip the text got crazy again. I have to power cycle it everytime. I have 2 SFIX chips and they both do the same.


Any Ideas?
thanks
 
BIG BEAR

BIG BEAR

SHOCKbox Developer,
20 Year Member
Joined
Dec 14, 2001
Posts
8,383
There has to be a bad connection hidden somewhere.Trace back to the cartridge connector pins
BB
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
The games play flawlessly, and in game text is perfect.
Except for the sound. No sound at all. But I didn´t looked about sound yet. First I wanted to fix the text. Only the text on bios messages is like in the above images.

Maybe I should look around Neo-IO?
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Thanks!

I disassembled my good working board to check the signals around NEO-I0, and I think I´ve found something, but I don´t know if it relates to the text issue:

Pins 19 (ORI0), 20(ORI1) and 21 (ORO0) are different between the boards. Pin 21 is SM1 Chip select, and in the working board is pulsing, but in the faulty one is stuck Hi, So SM1 is not being selected for anything. Thus probably the message of being unresponsive.

Both inputs on the other pins are: pin 19 stuck low on both boards, and pin 20 is pulsing on the good one, but stuck Hi on the faulty. But this pin 20 shoud be an input to NEO-I0 but its connected to the 1A input in the OR gate at the 74LS32 logic IC. Is it really connected between 2 inputs at different ICs or am I crazy too?

Could the SM1 justify anything about the text? I think this is only the reason for not having sound on the board…
 
F

ftadeufs

n00b
Joined
Mar 18, 2019
Posts
26
Hello guys,

I have set this board apart for a few months, and for my surprise when I powered it on, it showed perfect text! But after 3 or 4 seconds it started glitching very slowly from left to right. I managed to get a picture after it started that shows half screen good and half glitched. after some time everything was as before, and rebooting it doesn´t make the glitch go away... Anyone have any idea of what could cause this?

Neo Geo MV1-1 problems.jpg

thanks!
 
You must log in or register to reply here.
Top