Display Corrector update...

Greetings everyone.

In between building Sega Multigame boards (damn that gets dull after a
couple hours) I worked on the display corrector board. For those of you
that want to skip my "usual" technical babble-- the short version of the
story is this: got it working with Star Wars now and the results look
REALLY NICE!

Long technical-babble version (with some Star Wars troubleshooting
thrown in):

For some reason the display corrector as implemented for Major Havoc
didn't work worth a damn with Star Wars. After some poking and prodding
I came to the conclusion that the voltages present at the output of the
current->voltage convertor in the AVG on Star Wars are much smaller than
those on Major Havoc (or Space Duel, or anything else for that matter).
This was/is particularly mystifying since the component loads and
schematics are IDENTICAL. I'm writing this one off to programming
differences. It's as if Star Wars was written with the "scale"
intentionally turned down to give some dynamic range for the death star
explosion at the end of something. Weird.

Once I figured that out, I set about making changes to the display
corrector to "fix" the problem. I started by simply figuring that if I
just boosted the XREF and YREF voltages going to the DACs that the rest
would work. So, I put a 4.7K resistor in parallel with each 7.5K 1%
resistor off the collector of the 2n3906 in the linear scaling circuit.
This made the image larger, and everything seemed fine, except... That
broke scaling. This was odd, since the only time it was noticeable was
when "flying down to" the deathstar and the "star wars" scroller in the
demo mode. Everything else was fine.

I also managed to toast the 2N3906 in the VREF circuit (put the 4.7K
resistor across the resistor controlling base current to the 2n3906 by
accident. Oops.). This results in a vector display that looks as
though all the objects are "blown apart" slightly and listing down and
right.

My board that had the "Z intensity problem" (aka, all color guns were
stuck "ON") ended up having two problems-- a) a cheap socket for the DAC
had an internally open pin which was preventing good Z-intensity
infomation, and b) a TINY short around the intensity DAC's TL082 was
lifting the output rail high.

Anyway, all those aside, I came up with a good "corrector" circuit with
a variable "bow" control. I ended up removing the XBIP control inside
the display corrector. It only functions in conjunction with the VDR
(which is probably impossible to buy now-a-days anyway) and as near as I
can tell the VDR makes NO visible changes to the display. I remember
someone thinking it helped control "blooming" near the screen edges?

In the "neat" category-- if you omit (or bypass) the 100ohm series
resistor in the X-axis output of the AVG you get a really wild "ringing"
effect on the output. The neat part is that all "regular brightness"
objects are drawn w/out problems on the screen, but bright objects
(x-wing fire, TIE fireballs, etc) get a "wavy" appearance-- like
applying a varying frequency sine-wave to each individual vector. Looks
kinda cool, actually...

The display corrector looks great with Star Wars. I'll try it with
Major Havoc tonight. It has a disable switch to turn it off if using an
already "corrected" game. Right now it *really* looks like the right
way to do it is to take the 'inputs' right off the digital switches
(those two little jumper wires coming off the AVG on Major Havoc and
going to the card-edge connector? You got it... That's what those are
for.). I'll try it with the "monitor level" signals attenuated down,
but I think there might be some video quality compromises there. (Or
Atari probably would have done it instead of hand-modifying the Major
Havocs...)

-Clay

Clayton N. Cowgill Engineering Manager
-------------------------------------------------
/\ Diamond Multimedia System, Inc.
\/ Communications Division
Received on Mon Feb 2 15:08:17 1998