Atari Championship Sprint - Top 8 lines of sprites

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Note: This started out as a tech post here

Note 2: Thanks to ColinD mainly, he'd given me notes about a month ago when we saw this issue start up at the arcade but I was trying to understand what was happening.

Here we have one Atari System 2 Championship Sprint PCB with the following fault:



All the rest of the board is fine, the car sprites are doubling up the top 8 lines of them - displaying the full width no problem just repeating the top part.

After some poking around I was getting a bit lost, Colin chimed in with the help in the original tech post which resulted in me looking at the LS283 adder at 5M and 5L, both had what looked like signals going in and out - nothing obviously stuck.

Logic comparator was thrown on and immediately it disagreed with pins 9, 10 and 13.

Piggybacked a good 283 on the 5M and got this:



So we've got improvements. Pulled and replaced the 5M and now we have this:



So, another board lives.

The best I can understand of what's going on is that MOV0-8 going into 5M/5L are the vertical position of a given sprite, the 1V-256V is the current raster position.

These form an offset which via the pre-stored decoding in the PROM at 5P equates to adding and subtracting to get a data offset within the motion object ROMs

This happens for each line of the screen and where a sprite is within that raster line the current raster line taken from the sprites top line position is the offset plus the object number * 16 (16 lines per sprite) within the ROM is the final address for the current line to draw.

This data is then pushed into the motion object line buffers and offset using the shift controls via the 194's to put it at a definitive pixel X position along a line.

That line is then muxed with all the other stuff via the prioritising logic and ends up on the screen.

Yes, I only partly understand that so anyone who wants to explain the chain better please go right ahead as i'd love a better understanding of it.

Going back to my original post, the following I think are correct:

MOV0-8 are the vertical position of the given sprite.

MOPIC0-8 is the sprite number

MOSIZ0-2 are a sizing value but I don't think championship sprint uses these.

Anyway, thought i'd share the repair as a proper fix log so that its filed in the correct place. Those who follow me on YouTube etc expect a video to appear in the next day or two.

RGP2015-05-19 00:43:29
 

Hurray Banana

Moderator
Staff member
vacBacker
Feedback
8 (100%)
Credits
2,708CR
This is great James, you are really starting to get deep into this problem tracing work (do you actually play any games anymore
smiley1.gif
). Top job and great input from ColinD.
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
Thanks Eric, i actually play the games a bit more now than i used to but i love this side of it.

It totally fascinates me how they came up with ways to do things back then versus how we would do it now.

All games are no more than a highly specialised microcomputer with custom sound and graphics meant for one purpose at the time - now with us homebrewing as well, they have potential multiple purposes.

Next up... Paperboy with an interesting screen issue.......
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
frothmeister said:
Nice work James (and ColinD )
smiley20.gif


Keep it up, then I'll know where to send my Super Sprint and Paperboy boards when they innevitably break down again probably next time i power the cab on.
smiley36.gif

Fixed
smiley1.gif


They're not that unreliable although i did proceed with caution while desoldering the chip as i'm told they're a multilayer pcb.
 

frothmeister

Active member
vacBacker
Feedback
55 (100%)
Credits
868CR
Don't say that James, that you might curse me.
smiley15.gif


To be fair my Super Sprint has only ever had one breakdown in the 11 years I've had it and that was just that it lost a colour, think it was green.

I think P-man fixed it in his sleep tbh about 7 years ago and it's been rock solid since.

Paperboy has had 2 breakdowns, both in quick succession, maybe 4/5 years ago. P-man sorted both of them out too - think they were just down to bad RAM.

frothmeister2015-05-19 21:32:16
 

strykr

Active member
vacBacker
Feedback
11 (100%)
Credits
1,337CR
RGP said:
Experts?!? Dunno about that, i'm a dumb mo-fo just poking around blindly with a scope.

smiley4.gif

lol, im a dumb mofo plugging it in and it doesnt work and it goes in the "fix when i win the lottery pile"
smiley36.gif
.

Touch wood my champ sprint has held up so far, my super sprint crashes after about 20 seconds tho
smiley5.gif
.

I still love the fact that the pcb stack is 5 foot tall and today is about 20kb in a rom file
smiley36.gif
 

RGP

Meeter & Greeter
Feedback
5 (100%)
Credits
2,039CR
strykr said:
.... snip ..... my super sprint crashes after about 20 seconds tho
smiley5.gif
.

Interesting.....
smiley3.gif


Curious enough to take a look sometime in the future if you want. I'm always intrigued by the "resets after X time" problem as this reminds me of some software from years back.
 

karlcdoe

Active member
vacBacker
Feedback
7 (100%)
Credits
864CR
Depends on when it crashes out. I had a Capcom board that always crashed out during the same section of game/code and it was a dud program rom, programmed a new one and it was fine. If by crashing it is resetting then the next stage would be to find the watchdog disable (most atari pcbs of this era have them) and see if it locks up. then you have the usual yawn fests of thermal problems although 20 seconds is pretty quick.
 
Top