MV2F severe graphical issues

[neSneSgB]

I am working on a MV2F that started off doing nothing. I replaced the LSPC2-A2 to restore sync, and then the NEO-I0 after that to actually get an image on screen. With the Neo Diagnostics BIOS, the text is unreadable as it is either the wrong characters or has dots and lines all over.

On the UniBIOS, the crosshatch is there but is also messed up as it has the wrong tiles and characters selected. I noticed that it seems to be the lowest bit stuck, as R becomes S, H becomes I, F becomes G, etc.

With a game inserted, the NEO GEO logo is just white lines, there is no sound, and the game itself (Neo Turf Masters) is little more than white lines and blocks. This board had some minor battery leakage on it, but the traces near the NEO-ZMC2 and NEO-257s I checked first were all good.

I'm unsure where to look next. I swapped the NEO-G0 on a whim to no avail. I feel like this board might have been overvolted at some point considering both of those custom chips had faults, so I wouldn't be surprised if I had to replace more of those.

Any ideas would be appreciated.

 

[ack]

Based on the pictures it looks like data line 0 is stuck high or floating on the lower byte slow vram.

https://wiki.neogeodev.org/index.php?title=LSPC2-A2

Which should be pin 41 in the above pinout picture. Given you swapped the LSPC2-A2 I would double check that pin is solder ok and no bridges to adjacent pins.

It is odd the diag bios doesn't complain.

 

[ack]

I was thinking about this some more last night. Since the diag bios isn't detecting any memory issue with the slow vram. I think it may actually be an address line issue with the sfix.rom

 

[neSneSgB]

I pulled the SFIX ROM real quick and it verified fine. I also gave those pins on the LSPC2-A2 a reflow and it did not make a difference. Another thing I noticed is that the Neo Diag's look is not consistent, sometimes the lines go away but the characters are missing pixels. Also, a green haze starts to creep its way towards the right of the screen until the whole background is green. I did not notice any of this occuring with the UniBIOS's crosshatch.

 

[ack]

I was able to reproduce your initial images by forcing pin 7 on the SFIX rom to be high. I would see what the pin is doing with a logic probe, it should be pulsing.

The green haze effect can be an indication of the floating pin/line. When its happening I would try applying downward pressure at different locations on the board and see if it has an effect. If its just a cold solder joint doing that can help narrow down where on the board the issue is. I would also suggest logic probing the palette ram and the 2x74LS273 near the jamma edge to see if anything is floating.

 

[neSneSgB]

Here's a video I recorded of the issue, you should be able to hear my logic probe change in tune with the screen. It starts off constantly high, toggles a little bit, then goes high again, toggles a little and goes low, etc.

View: https://youtu.be/CkPSpEodY2w

In this video, the green haze was actually dark blue, but it's not visible on the video; it just looks like nothing is there.
I tried to press on the board prior, but it didn't seem to have any effect on the screen, as it randomly changes by itself already.

I forgot to mention this, but I had replaced the 273s prior to getting video as they were starting to corrode. I also replaced the palette RAMs at the same time as I wasn't sure what was causing no video. I will still probe those locations to see if anything looks awry.

Is there a chance the Neo-I0 I replaced is faulty? I don't recall the Neo I pulled it from having this issue, but I do have another chip I could borrow to see if it fixes.

 

[ack]

Seems like you are trying to resurrect a parts board based on the number chips you've already replaced.

I would check pins 2-5 on the NEO-I0 and see if any are floating.

 

[neSneSgB]

Yeah I don't know what happened to this board, but considering the I0 and LSPC2 were faulty, maybe overvolted? I don't think the palette RAMs were faulty, but I just replaced them since I have so many 62256s. The board I grabbed these chips from is more of a parts board due to the corrosion damage, while this one is actually pretty good near where the battery resides.

I still worry that I made a bridge on the LSPC2 with those extra fine pins, but it doesn't sound like that is causing the issue at hand.

I also have another MV2F that looks like the picture attached, so somehow the one I'm working on is the best choice to revive out of these three.

 

[neSneSgB]

I gave the NEO-I0 a reflow and thankfully the incorrect text issue cleaned up, so I must have had a solder bridge somewhere. Now the text reads perfectly, but still loses some of its pixels as the green haze begins.

 

[neSneSgB]

I was also able to see signs of Neo Turf Masters working, so that's a good thing too.