MV-2 Flickering Pixels on edges

mikejmoffitt

Mickey's Coach
Joined
Feb 6, 2014
Posts
578
I've had this MV-2 for a long time, and it had vertical lines and very corrupt looking flickering graphics. Just a few days ago, I got around to addressing the problem - the old battery corrosion had destroyed almost all of the traces coming down from NEO-ZMC2, which does pixel multiplexing. The graphics are now almost perfect. I also repaired one bad ADPCM-A address trace to fix sound.

I am left with only one problem. I've made sure my wiring from my repairs is correct and there are no shorts or egregiously long wires. The problem is that occasionally the edges of sprites will have flickering pixels that go between two states. This is most evident when H-shrinking is being applied but not always. I've tried to capture some evidence of it in a video:

http://csh.rit.edu/~moffitt/img/mslug2.mov

There are a few vias I can push on with my finger and make the dots change a little bit - something is definitely floating when it should be driven. The problem is I've followed the traces around and they are hard to follow as they go closer to the rest of the graphics chipset. I suspect the other end connects to ZMC2 but as they go under ZMC2 I can't really tell what is supposed to go where. If I can avoid de-soldering ZMC2 that would be great.

If anyone has any pointers I would appreciate it. I'm so close to having this as a fully working board.
 

skate323k137

Professional College Dropout
10 Year Member
Joined
Jan 7, 2013
Posts
4,186
Yep, I can see it :)

Sadly I'm probably of no help on this one, but best of luck.
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
Looks like a connection from the ZMC2 to NEO-B1, iirc.
 

mikejmoffitt

Mickey's Coach
Joined
Feb 6, 2014
Posts
578
Unfortunately this is an MV-2 which has no NEO-B1 - just PRO-C0, which does connect to ZMC2's multiplexed pixels (both of them). All 8 connections look fine for that. It's correct so much of the time, and it only recurs on certain graphics tiles. I wonder if something is wrong with C rom addressing...
 

Xian Xi

JammaNationX,
15 Year Member
Joined
Dec 1, 2005
Posts
27,748
Then it would be the B0 and/or LSPC0. I had one that had that chipset that had that problem. I thought you had the MV2F so I tried to see which one it could be. I don't remember exactly what it was though.
 

mikejmoffitt

Mickey's Coach
Joined
Feb 6, 2014
Posts
578
With closer inspection, I've noticed it only seems to happen when sprites are placed on an odd H-position. It might be related to the way ZMC2 serializes pixel data two at a time - pixel A and pixel B. I spent a few hours probing about trying to figure out what should go where but we don't have schematics or pin info for this generation of the chipset so it's a bit harder than I wish it was...
 

mikejmoffitt

Mickey's Coach
Joined
Feb 6, 2014
Posts
578
Yes, that's what I'm referring to - all of the graphics chipset I have are in the "to-do" section.
 

mikejmoffitt

Mickey's Coach
Joined
Feb 6, 2014
Posts
578
Changing the supply voltage makes the glitchy pixels move to other areas of the screen but I wouldn't say it really helped. Something is still definitely wrong.
 

daskrabs

Ace Ghost Pilot
10 Year Member
Joined
Aug 23, 2010
Posts
1,313
I'm seeing this on my 4-slot. Did you ever resolve?
 

mikejmoffitt

Mickey's Coach
Joined
Feb 6, 2014
Posts
578
The fix was to remove the previous owner's "fix", which ran a jumper wire from some random point A to random point B.
 
Top