Wonder Boy - minor sprite issue

[big10p]

I have an original Wonder Boy PCB which has a minor, but annoying, sprite issue. The pixels on the far right of the sprite seem to change size slightly, as they scroll across the screen, giving them a slight shimmering effect when playing...

Notice the black pixels on the far right of the rock and apple. First pic shows how they should look, second pic shows the pixels sort of half the width they should be. It changes between the two as sprites scroll across the screen, giving the shimmering effect. I've checked in MAME, and they don't do this. I've SLICEd the sprite RAM, which seems OK.

Any idea what can cause this? I can't even comprehend how half-width pixels can even be rendered.

 

[Lurch666]

Are you playing on an LCD or CRT?
if an LCD it could be your screen and not the PCB causing the issue.

 

[big10p]

Are you playing on an LCD or CRT?
if an LCD it could be your screen and not the PCB causing the issue.

Yeah, that's what I thought at first, but it does it on the CRT in my cab, too. Also, MAME doesn't do it, even on an LCD.

 

[big10p]

Here's a short vid that shows the issue more clearly, on the snail sprite.

New video · Tuesday, Apr 14 🎬

Tap to view!

photos.app.goo.gl

Seems to happen to only the last pixel on each scan line of the sprite, and only on alternating sprite X coordinates (either all even coords, or all odd coords).

 

[TheDaddy]

Crikey how did you even notice . Thats gonna be a hard one to find me thinks ! , Thats if I am even seeing what you are.

Dave.

 

[big10p]

Crikey how did you even notice . Thats gonna be a hard one to find me thinks ! , Thats if I am even seeing what you are.

Dave.

The snail is slow so shows it more clearly. When movement is faster, it creates the shimmering effect on the edge of sprites.

 

[TheDaddy]

The snail is slow so shows it more clearly. When movement is faster, it creates the shimmering effect on the edge of sprites.

It like its not updating the screen fast enough ?? Wonder if its a simple as a ram issue , timing maybe ?? Not sure to be fair just pissing in the wind.

Dave.

 

[big10p]

It like its not updating the screen fast enough ?? Wonder if its a simple as a ram issue , timing maybe ?? Not sure to be fair just pissing in the wind.

Dave.

Yeah, this is along the same lines as my thinking, but I don't know enough about this stuff yet to know exactly what. I even tried faster sprite ROMs (as I'm not sure the installed ones were originals). The sprite buffer ram if Fujitsu, but it seems to be working ok using SLICE.

 

[Georgian2]

It like its not updating the screen fast enough ?? Wonder if its a simple as a ram issue , timing maybe ?? Not sure to be fair just pissing in the wind.

Dave.

There was once an Wonderboy for sale here on the forum where everything moved in slow motion. I wonder if this issue is the beginning of that Problem.
An oscilloscope would be better here to see the signal levels. Something weird is going on.

 

[big10p]

There was once an Wonderboy for sale here on the forum where everything moved in slow motion. I wonder if this issue is the beginning of that Problem.
An oscilloscope would be better here to see the signal levels. Something weird is going on.

I've got a scope but unsure which signals I should check. Maybe I should probe the sprite ram to see if something looks off.

Having a schem would help, but not available it seems. Also, this board has a couple of daughter boards, so probing ICs underneath is difficult, to say the least.

 

[Georgian2]

There are schematics for one system1 board. Not sure now which model. Thye are not the same but very close.
http://techdocs.exodusemulator.com/Arcade/SegaSystem1/index.html

 

[Jacmar]

That is a real specific weird fault, not sure how you'd see it on a scope but certainly won't help to look.
Its like a very marginal timing issue, maybe a counter is a tiny bit off or the reset/clock for it is a tiny bit out .... dunno like @TheDaddy says .. pissing in the wind a bit. Will be a tough one to figure out ... I'd be looking at the chips after the ROMS, how does the data turn into single pixel data on the screen? shift registers ? check them .... can you pop out the sprite roms one at a time, see if the error stays or goes away.... might help narrow it down ..

 

[pubjoe]

I've seen this exact thing before. I can't remember if it was with emulation or a PCB, but it might help if I can find it. It looks minor in the photos but it's really clear in gameplay when every sprite is doing it. Annoyed me too.

I'll have a look.

 

[big10p]

There are schematics for one system1 board. Not sure now which model. Thye are not the same but very close.
http://techdocs.exodusemulator.com/Arcade/SegaSystem1/index.html

That could me useful. Thanks.

 

[big10p]

That is a real specific weird fault, not sure how you'd see it on a scope but certainly won't help to look.
Its like a very marginal timing issue, maybe a counter is a tiny bit off or the reset/clock for it is a tiny bit out .... dunno like @TheDaddy says .. pissing in the wind a bit. Will be a tough one to figure out ... I'd be looking at the chips after the ROMS, how does the data turn into single pixel data on the screen? shift registers ? check them .... can you pop out the sprite roms one at a time, see if the error stays or goes away.... might help narrow it down ..

It's also weird that it only does it on alternating sprite X coords. I'll try removing roms one at a time. I didn't even know it would work with a ROM missing. Lol

 

[big10p]

Ok, I tried removing the sprite ROMs one at a time, but it didn't have any effect on the issue.

I've just seen a thread on klov about the same PCB as mine (834-5191-01) and apparently the Up n Down schems are the same.

Wonder Boy (Original, Sega 834-5191-01 PCB) factory jumper wire mod?

I recently got my Wonder Boy PCB mostly working (it crashes at random, but I think this is due to corroded sockets which I plan to clean or replace), but I am unsure of where exactly this green wire is supposed to be soldered. One end is connected to the Sega 315-5012 chip on pin 6, the other is...

forums.arcade-museum.com

 

[Jacmar]

trace path of data lines from sprite roms up to where they mix with other gfx layers, Slice those IC's (latches & shift registers) .. Also check clock inputs to those

 

[pubjoe]

I finally found a ten year old video that I thought might have demonstrated this but no. Cute video though

New video · Thursday, Aug 18, 2022 🎬

Tap to view!

photos.google.com

I've definitely seen this effect. Would have either been Groovymame (perhaps it was a particular romset) or my PCB. I found some pics of a board I sold late-2016. Bit of a long shot but you haven't ended up with it have you? No. 159157. I was informed it was a conversion. I had no idea of this until posting it for sale but maybe that was the cause of the glitch.

 

[big10p]

I finally found a ten year old video that I thought might have demonstrated this but no. Cute video though

New video · Thursday, Aug 18, 2022 🎬

Tap to view!

photos.google.com

I've definitely seen this effect. Would have either been Groovymame (perhaps it was a particular romset) or my PCB. I found some pics of a board I sold late-2016. Bit of a long shot but you haven't ended up with it have you? No. 159157. I was informed it was a conversion. I had no idea of this until posting it for sale but maybe that was the cause of the glitch.

View attachment 56775
View attachment 56774

Haha, cute vid. No, that's not my board. Mine is a different layout...

It's got the same board number as the one in the klov thread, but still has a few differences to that one. There seems to be so many versions! Maybe it's an issue with my specific version, I don't know. 🤷‍♂️

 

[TheDaddy]

I got 3 of them !! None working though or I could test to make sure its not a factory upgrade fault. Very possible and people just not noticed maybe ?

Dave.