RE: Tech:Asteroids

From: Clay Cowgill <ClayC_at_diamondmm.com>
Date: Fri Feb 12 1999 - 12:42:05 EST

> Anyone ever work on an asteroids board that worked fine, passed all
> self
> tests, plays
> ok. But, in test mode the cross hatch pattern in the background is
> missing,
> and in play mode
> the customer stated some vectors were still being drawn when it
> shouldn't,
> (Monitor
> brightness was ok), couldn't tell what was missing/there myself on the
> scope, looked ok to me.
> I guess its time to build up the bw xy monitor project that I keep
> putting
> off...
>
The brightness of the crosshatch is set lower than most objects in the
game. It's possible the monitor brightness was just too low--
particularly if it looks OK on a scope.

As a long shot-- the cross hatch and all the diagnostic screen stuff
lives in the very beginning of the vector ROM. It takes up a LOT of
space-- I don't remember if it's 128 or 256 bytes. If you have a bad
vector generator instruction in the display list for the cross hatch it
wouldn't display all of it. I stole some space for a replacement "0"
shape for the color Asteroids mod by forcing an RTS early in the test
mode display and taking back some bytes...

On the Really Bizarre list: Asteroids seems to have some kind of spiffy
error detection in the vector ROM test area. Take any byte in the
display list for the test-pattern of the VROM (the first 128 or 256
bytes or so) and change it to ANYTHING. Fire up the game in test mode.
The test routine will show the *original value* of the byte you changed!
Not just a "location xx bad" but "AA should be BB". Works on any byte.
Almost like they have a "spare" table stored somewhere to compare it
to... Pretty neat.

> So, I've checked the outputs on that dac's expecting that maybe they
> were
> being clipped,
> but the dac outputs reflect the missing crosshatch pattern so I'm
> ruling out
> the dacs. Kinda
> got me stumped, I did notice that in DAX10 (load 10 output from the
> ls157
> or the state machine if I remember right) seems to be pulled low, but
> I've
> replaced the components that are driven and no go, Isolation of that
> line
> puts the game into watchdog as the state machine stops. So, I just
> assuming
> that I'm not seeing the output
> on my analog scope fast enough.
>
You'll not see much activity on the higher order DAC outputs unless
you've got a pretty fast scope... I was working on a board a couple of
nights ago and was hunting down bad outputs from the VG to the DACs. If
you've lost a high-order bit (like bit 10) the effect will be pretty
evident across the whole screen. In my case about three of those
!@#$!@#$ 74LS399's had bad outputs (one of which was bit 10 on X) and
the whole display was shoved into the "left half" of the screen. (It
must be a pretty old Asteroids board-- all the rest I have have the
399's redesigned with '157/'273 pairs instead.)

-Clay
Received on Fri Feb 12 11:42:29 1999

This archive was generated by hypermail 2.1.8 : Fri Aug 01 2003 - 00:30:52 EDT