MV-1AX Flashing Green Screen and Watchdog Clicking

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
:scratch:
Looking for guidance / help with a MV-1AX board. I repaired several bad traces, replaced the backup battery, checked the main 24Mhz Oscillator, checked the 32.768Khz XT, and still getting the flashing green screen with the watchdog reset/clicking sound. In the pictures:

Blue screen with garbled video is with all dip switches “on” (Ram test position)

Light Blue screen is with all dip switches off (normal operation)

Flashing Green is with cartrdge in slot and all dip switches off.

Hopefully someone can point me in the right direction as t my next steps with repairing this board, thanks in advance.
 

Attachments

  • A3492E8B-F5F2-4823-9D50-D55F85CF4105.jpeg
    A3492E8B-F5F2-4823-9D50-D55F85CF4105.jpeg
    1.1 MB · Views: 144
  • 350DE60A-3B6B-4230-A698-D610D1C77CA4.jpeg
    350DE60A-3B6B-4230-A698-D610D1C77CA4.jpeg
    935.3 KB · Views: 143
  • 74F8EDFD-DDB8-4110-8B72-E4FF964D7F6C.jpeg
    74F8EDFD-DDB8-4110-8B72-E4FF964D7F6C.jpeg
    1.7 MB · Views: 147
  • BA531820-E955-476F-91FF-D78B0BBA2597.jpeg
    BA531820-E955-476F-91FF-D78B0BBA2597.jpeg
    1.3 MB · Views: 146
  • 70A6C014-D033-414E-8341-339B87EA0C57.jpeg
    70A6C014-D033-414E-8341-339B87EA0C57.jpeg
    1.4 MB · Views: 139

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
Update:

- Checked the traces on the HC32 in the Backup area and they all have good continuity (approx. .20 ohms)
- Checked the connections between the 4990 and Neo-F0 and they are good
- Checked the connection between the 4990 and HC32
- Checked the connection between the 4990 and the 32.768 XT
- As a precaution changed the 4990 and HC32 , with replacements from a donor board
- Checked the traces / connections on the backup RAM3 & RAM4 as well as the Main RAM1 & RAM2
- Just for Fun changed RAM1 / RAM2 / RAM3 / RAM4 with RAM from donor board
- Checked to make sure proper +5v VCC is being applied to Board
- Checked to make sure proper +12v is being applied for Audio amp (Which works very well, very loud)
- Checked VCC at both the 4990 and the HC32 in the Backup area, and they are at approx. +4.69v
- Checked a few of the connections that tie the Backup RAM3 & RAM4 back to the 68000

Still same issue, going to start putting alcohol in my coffee at this point, or get some sleep, have not decided yet.

Any How, as always any help is greatly appreciated, would love to get this system back to life.

:confused::scratch:
 

clarkindustry

New Challenger
Joined
Sep 15, 2005
Posts
59
strange, this problem born from oscillator of backup ram.
i remember some of my oldest fix.
have you tryed to replace it ?
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
strange, this problem born from oscillator of backup ram.
i remember some of my oldest fix.
have you tryed to replace it ?

I'll look at one of my donor boards to see if I can snag a spare 32.768Khz Oscillator and swap it in. Keeping my fingers crossed.
 

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
I assume you noticed the board is missing the PAL1/2 chips on the left side?
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
Good call, I suppose due to my lack of sleep I completely overlooked this. Next question is where to obtain replacements for the PAL 1 & PAL 2 Chips. I searched online and have trouble finding PAL20V8H-15, but did find PALCE20V8H-15 and GAL20V8H-15 as substitutions. However even though I was able to find the necessary JEDEC images to burn to the PAL Chips (Thanks to https://wiki.neogeodev.org/index.php?title=MV1A_PALs) I don't know how I would burn them as I am not finding too much on burning "PAL / GAL". Can anyone point me in the right direction for getting some replacements, and or how to burn them myself? Thanks in advance...
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
Just ordered some Replacements PALs to burn the JEDEC image onto, we’ll see how that goes.
 
Last edited:

GadgetUK

Ace Ghost Pilot
Joined
Sep 27, 2013
Posts
1,323
I would start by getting the SMK Dan diagnostic BIOS on there tbh. See if that gives you anything other than a watchdog. I've done a number of videos on these boards in the last week. Interesting the wires there on the Neo-Buf, seems to suggest trace damage from a swapped chip maybe, or perhaps someone was trying to bypass a failed IO pin.
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
I would start by getting the SMK Dan diagnostic BIOS on there tbh. See if that gives you anything other than a watchdog. I've done a number of videos on these boards in the last week. Interesting the wires there on the Neo-Buf, seems to suggest trace damage from a swapped chip maybe, or perhaps someone was trying to bypass a failed IO pin.

I put the Diagnostic Bios on the board and still getting the watchdog, however I have yet to install/replace the missing PAL chips. Regarding the wiring in the picture, that is my patch work for the broken traces that are on the board. Hoping to get the PAL chips soon, and then to figure out how to program them with the correct JDEC. Hope my EEPROM programmer will do the trick.

As a side note your video on the Z80 from hell has given me ideas to repair my own Z80 from hell, as I posted on your YouTube video. Thanks Again.
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
Just received the PALCE20V8H-15 I ordered to replace the missing SNK1A & SNK1B on the MV-1AX board. I have also obtained the JEDEC images for them (thanks to https://wiki.neogeodev.org/index.php?title=MV1A_PALs). My Next dilemma is how to burn the JEDEC images onto the PAL chips. I have been doing some reading and see that the PAL20V8 is compatible with the GAL20V8, which I though I could use to my advantage as my programmer (TL866a) shows the GAL20V8 as an option. However this did not work, I also have a PICKIT 3 but that doesn't handle anything other than Microchip devices. Does anyone have any suggestions on how I can get these chips burned? Or can someone burn them for me?

Unfortunately I can't further test/repair the board until I get them installed. From the NEO-GEO Dev site "The MV1A MVS boards have two [PAL] PAL20V8H-15 chips which are used to provide some of NEO-C1's functions. If any of them fails, the board certainly won't start up. SNK1A does address decoding, SNK1B generates the DTACK signal."

Thanks in advance all, for any help or guidance you can provide.

:scratch:
 

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
I can burn a pair of them for you.
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
Ok, so just installed the missing PAL ICs SNK1A & SNK1B (Thanks ack! for sending them to me)

- Still getting Watchdog click, however the image on the screen is now green with garbage pixels, instead of Blue with garbage pixels as it was in the image I posted originally

- Checked again for broken traces, found 1 and repaired it, still no change.

- As a precaution changed RAM 5 & RAM 6, as I had some spares and those 2 had not been changed when I changed RAM 1 through RAM 4

- Probed the logic levels on the HC32 and D4990a
(Levels H= High, L=Low, P=Pulse, NC=No Connection)

HC32 (Quad OR Gate) Readings seem good
1 through 6 - H
7- L
8 & 9 - H
10 - L
11 through 14 - H

D4990a
1 ~ 3 - H
4 - NC
5 - L
6 - H
7 ~ 9 - L
10 - P
11 - L
12 - H
13 - NC
14 & 15 - Brief High, then pulse, then logic probe shows no reading (could be due to the fact these are the connections to the 32.768Khz XT)
16-H

- Cleaned connections around the board and the ICs with some Isopropyl 99%, and checked solder connections, no change.

Any thoughts or ideas gentlemen? Would anyone by chance have a MV-1AX that can give me some logic values to compare? as always thanks in advance. :scratch::oh_no::scratch:
 

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
What bios are you using? Sounds like a stock one if its a green screen with junk. what happens with diag or unibios?

I have an MV1AX I can check the probes with later today.

I would also suggest looking over the bottom of the board and make sure no components are damaged or missing.
 

GadgetUK

Ace Ghost Pilot
Joined
Sep 27, 2013
Posts
1,323
I've got 9 of these I am working on atm - one of the Neo-Buf chips is what I would place my bets on tbh. Out of 9 boards, almost every one of them has MULTIPLE neo-buf failures, and on 3 boards they've caused watchdog.
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
What bios are you using? Sounds like a stock one if its a green screen with junk. what happens with diag or unibios?

I have an MV1AX I can check the probes with later today.

I would also suggest looking over the bottom of the board and make sure no components are damaged or missing.

- With Both Uni-Bios 3.3 and Diag Bios getting same condition.

- Checked for missing components, replaced 2 100nF Caps that looked iffy.
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
I've got 9 of these I am working on atm - one of the Neo-Buf chips is what I would place my bets on tbh. Out of 9 boards, almost every one of them has MULTIPLE neo-buf failures, and on 3 boards they've caused watchdog.

Since the Neo-Buf chip is a dual Octal buffer, when testing it should I just compare the logic level on the "A" (A0 ~ A15) address and then look for the same state on the "B" (B0 ~ B15) address?
 

ack

Ninja Combat Warrior
15 Year Member
Joined
Apr 9, 2009
Posts
538
This is what my probes get with a stock bios, no cart sitting at crosshatch

Code:
HC32 (next to D4990a)
1  H/P
2  L
3  H
4  H/P
5  L
6  H
7  L
8  L
9  L
10 L
11 H
12 H
13 H
14 H

D4990a
1  H
2  H
3  H
4  NC
5  L
6  H
7  L
8  L
9  L
10 L
11 P
12 H
13 NC
14 
15 
16 H
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
This is what my probes get with a stock bios, no cart sitting at crosshatch

Code:
HC32 (next to D4990a)
1  H/P
2  L
3  H
4  H/P
5  L
6  H
7  L
8  L
9  L
10 L
11 H
12 H
13 H
14 H

D4990a
1  H
2  H
3  H
4  NC
5  L
6  H
7  L
8  L
9  L
10 L
11 P
12 H
13 NC
14 
15 
16 H

Thanks ack for the Logic Values, I am comparing them with the board I have using the same configuration (Stock Bios / No Cart).

Some take away observations I am making regarding the HC32 (next to D4990a). Since this is a Quad 2 input OR gate, as long as there is a pulse or high condition, there will be a High Output.

Pins 3 , 6, 8, 11 are the OR gate outputs. My readings all agree (output wise) with yours except for Pin # 8 where I am getting a High output due to OR Gate logic from the combination of input at Pin # 9 High and Pin # 10 Low. On your good board measurement your pin # 8 is Low and both the inputs #9 & #10 are Low as well. I think one of the things I should do is trace back where my HC32's Pin #9 is being fed from as it is coming in High.

Also GadgetUK mentioned Checking the Neo-BUF chips as well, trying to determine the best course of action to test them. As I have 2 MV-1AX Boards with similar conditions so I can't beside which IC's are functioning correctly at this time.
 

GadgetUK

Ace Ghost Pilot
Joined
Sep 27, 2013
Posts
1,323
Since the Neo-Buf chip is a dual Octal buffer, when testing it should I just compare the logic level on the "A" (A0 ~ A15) address and then look for the same state on the "B" (B0 ~ B15) address?

The OE enable needs to be low, and the direction pin dictates which way the data goes, but in theory I think if you check when OE low they should match. I've got numerous 1A and 1AX boards here that watchdog (even with diag BIOS) because they are outputting when they shouldn't - randomly.
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
The OE enable needs to be low, and the direction pin dictates which way the data goes, but in theory I think if you check when OE low they should match. I've got numerous 1A and 1AX boards here that watchdog (even with diag BIOS) because they are outputting when they shouldn't - randomly.

Thanks for the information, I decided to check the connections to each of the Pins on the NEO-BUF a little more carefully and found something interesting. The NEO-BUF (GA9) has short between the VCC (Pins 19 & 43) and Ground (Pins 17, 18, 41, 42) of 14.5 ohms. This is definitely not good, so thinking this to be the issue I removed the NEO-BUF and rechecked the contacts, the VCC short to ground still exists.

Just to make sure I was not crazy I checked my other MV-1AX board and there is no VCC short to ground, and the average resistance between VCC and Ground and the main board test points near the Jamma Connector is about 480 ohms (and this is also the same on the MV-1FZ & MV-1FZS).

*** All testing for the short circuit is being done without power, BIOS removed, SNK1A & SNK1B removed, No Battery, Jamma Unplugged, just the bare board ***

- Not sure which component could be causing the short to ground condition, I removed the other 2 NEO BUF Chips (GA7 & GA8), still getting the 14.5 ohm VCC short to ground.

- Removed the 74HC273 (U11) & LS232 (U12) still getting 14.5 Ohm VCC Short to ground.

- Tried visually inspecting the board and ICs for any clues as to what could be causing the short, no luck.

- Even though the SMD ceramic Caps rarely go bad, and my ESR meter shows them as good. I decided to remove PC32, PC33, PC35, PC42 as they connect to the ground plane and are located in the same area as the NEO-BUFs, still no luck.

- Aside from continuing to remove components, trying to determine if there is a better way to find this short circuit.

:scratch::scratch::scratch::scratch:
 

Dropwire

Kuroko's Training Dummy
Joined
Jul 27, 2018
Posts
75
Still trying to track down the short circuit.

The Neo-MGA/T is showing a minor short to ground Pin 93(VCC). Removed the NEO-MGA and see a slight improvement the short to ground is now approx 35 ohms instead of 14.5 ohms.

Instead of randomly trying to remove more components, I have tried the laser temperature method of checking component temperature with the idea the “shorted” component will have a higher temp. So far no progress.

Also trying the DMM Resistance test using my Fluke 45 as it has the lowest resistance measure of the meters I have. The idea is measuring between board ground and the various components that tie to ground looking for the lowest resistance.

Any thoughts or ideas gents? This board is becoming the bane of my existence.
 
Top