Neo geo MV2F garbled screen

[makedome]

Hello, I'm repairing this 2 slot neogeo board that shows garbled screen. No click of death because halt and reset signals are high.
It had some corrosion around battery but tested all traces and are ok.
Palette rams 43256 next to lspc2 and hi vrams 4364 have we high not toggling.
Also noticed that the small oscillator 32.768khz doesn't have any signal. I'm not sure my oscilloscope could read it's signal correctly as it did for the other 24 MHz one. The NEC 4990 dip14 does not show anything on pins 12 and 13 where the scillator is connected. So I would replace it. Any other hints?

 

[Kujako]

Have you checked continuity on the SP1 BIOS to the CPU?

 

[ack]

If the screen is static, I would verify you are giving the board enough voltage.

 

[makedome]

hello. Voltage is ok. About the continuity between sp1 and cpu, do you have schematics? Btw data lines are all connected to sp1 and address lines to neo-eo chip

 

[Kujako]

hello. Voltage is ok. About the continuity between sp1 and cpu, do you have schematics? Btw data lines are all connected to sp1 and address lines to neo-eo chip

I'm not aware of a pin mapping for the EPROM BIOS chips to the 68k CPU. I've got this one for a MV1C, but for other systems I just compare the pinouts.

You could also check some of the schematics. https://wiki.neogeodev.org/images/c/cc/Mv1fs-page1.jpg pinouts from SP1 to the CPU should be the same for the most part.

 

[nauman]

No click of death because halt and reset signals are high.

At a normal start up, the reset signal would be low for a short period of time and then keep it low until watchdog releases it. https://wiki.neogeodev.org/index.php?title=Reset
If there are no click-of-death afterwards (the reset line i changing often), then the system assumes the code is working (watchdog is getting kicked), or the watchdog is malfunction. All this reseting is happening before BIOS ROM is read and started. A missing reset-sequence could be the cause of the weird graphics.

Also noticed that the small oscillator 32.768khz doesn't have any signal. I'm not sure my oscilloscope could read it's signal correctly as it did for the other 24 MHz one. The NEC 4990 dip14 does not show anything on pins 12 and 13 where the scillator is connected. So I would replace it. Any other hints?

Make sure you are using a 10x probe or a 10M ohm resistor in between the probe and pad, this clock is annoyingly sensitive to additional resistans. A missing BRAM clock wouldn't cause the error you are seeing though, would more likely be a solid green screen/click of death scenario.

 

[makedome]

hello unfortunately this pinout seems to be totally different for mv2f. found some connections here https://www.neo-geo.com/forums/index.php?threads/mvs-z80-error.238470/#post-3711986. But the fact that the nec 4990 (14 pins not 16 as the picture) is not oscillating on pin 12 and 13 it's not a good sign.

 

[makedome]

At a normal start up, the reset signal would be low for a short period of time and then keep it low until watchdog releases it. https://wiki.neogeodev.org/index.php?title=Reset
If there are no click-of-death afterwards (the reset line i changing often), then the system assumes the code is working (watchdog is getting kicked), or the watchdog is malfunction. All this reseting is happening before BIOS ROM is read and started. A missing reset-sequence could be the cause of the weird graphics.


Make sure you are using a 10x probe or a 10M ohm resistor in between the probe and pad, this clock is annoyingly sensitive to additional resistans. A missing BRAM clock wouldn't cause the error you are seeing though, would more likely be a solid green screen/click of death scenario.

using the 10x on the oscilloscope probe gave me the clock on nec 4990.
For the reset signal it never starts low. It's always high. What should I search for the watchdog circuits?

Also for the click of death at the starting, I can hear a repeated high frequency noise for a couple of times and then I have that screen.
here you can find a video

View: https://youtu.be/-LOpsduEbpc


Thanks

 

[nauman]

using the 10x on the oscilloscope probe gave me the clock on nec 4990.
For the reset signal it never starts low. It's always high. What should I search for the watchdog circuits?

The watchdog / reset circuit is a core part of the B1 ic. If the reset doesn't happen, it can point to a broken B1, and maybe reset is not the only part that is actually broken inside it. The screen you are seeing comes from the LSPC still producing a picture, but with random data from the RAM. This is always happening and if you would start the system without a BIOS in the socket, this is what you would been seeing. Except that it would be a "rolling" screen due to the resets from the missing watchdog events.

Hard part is to really know if the B1 is the core reason for the system failing. Without swapping it, I think it is hard to really know :/

 

[ack]

Its possible watchdog has been disabled, J2 jumper above the NEO-B1 will do this if jumped.

I would check the NEO-D0 chip and make sure its generating the proper clocks and those clock are making it to the 68K, NEO-B1, LSPC2-A2.