TECH: Asteroids Deluxe board repair story/log
Dave
I acquired a Asteroids Deluxe board, and then a full upright.
<rant
The board is an example of what you get when a dealer lists the item
as 'unable to test'. Unwilling, rather, I suspect. I would note that
missing components (PROM, Crystal) were not indicated on the auction.
When you buy a dead board, you can assume that most of the time one
thing is wrong. Apparently there are dealers out there that don't
feel compelled to disclose a history of canibalization. Be warned,
ask questions, bid accordingly.
/rant>
I've got the board almost fully working (just in time, it turns out).
Here is my repair log so far. I'm posting it in case someone may find
it useful some day. At least it will be part of the google groups
database. If you have any insights, I'd be interested to hear.
** Received DOA from Ebay auction. It is different from the board that
came with my cab in little ways, like routing of some traces. For
example, all of the four bridge diodes (CR7-10) are supposed to
mount the same way, but CR8 is mounted opposite the others on the
other board.
** No crystal. No vector state machine PROM at C8. 24V input line is
vaporized. The bridge rectifier diodes are fried.
** Replaced crystal and C8 prom. Clock is OK. Repair 24V input trace
with a wire. Replace rectifier diodes.
** CPU constantly resetting. Lifting PIN 1 on 74LS42 at K5 still
gives resetting so it's the CPU section. Data is changing on ROM
and RAM address lines. Replace CPU. Now no longer resetting.
** hook up a little op-amp audio circuit to the board, hoping to
simulate blind play. Get series of beep tones indicating a bad
RAM. replace.
** Try in cabinet. No video. Voltages on 7XXX regulator outputs are
zero. Replace 7915, 7812, 7815, 7805. Replace the big 470uF
electrolytics (C72,C73). Now we have video. Thrust and explosion
sounds are present, but very weak. Game is playable. Self-test
reports 'ERROR'. The manual states that this means a bad POKEY.
On powerup, reset, or leaving test mode, the high-score screen
displays 'FFFFFF' and redraws 3 times before entering attract mode.
Subsequent high score screens are empty until you play a game, but
no high score is stored across power downs.
** Replaced the POKEY. Now all the sounds were audible but very weak.
** injected a 1v 1kHz sine wave into the audio Op-amp input (pin 9 of
P11, LM324), via a 1k resistor. The output at pins 8 was 1v and not
10v. Replaced the LM324. Sound is perfect now.
** At this point, game is playable. Ship motion and turning is
smooth. aliens are smooth. However, the tumbling asteroids appear
'jerky'. This is because every few degrees of rotation, an
asteroid is drawn inaccurately so that the ends don't meet. This
problem went away when I replaced the (original Atari) ROM set
(including the VECTOR ROM) with a fresh-burned set of the REV-2
EPROMs. However, the original ROM set works fine in the board I
got with my cab--the asteroids tumble smoothly. I wonder if the
vector EPROM access time is just slightly out of tolerance?
** At this point, the only remaining problem is the 'FFFFFF' high
score wierdness. I'm measuring -27.4V across C49. I tried
swapping the EAROM (N9) with a known good one from my other board,
without improvement. The EAROM write event is difficult to
replicate on the bench, although the 244 (M9) should get strobed
every reset. I suppose I'll just start replacing the support
logic, unless the 'FFFFFF' means something to somebody out there.
I'd be grateful for any suggestion.
P.S. The board that came with the cabinet now suddenly stopped
working. The problem is in the vector generator, since disabling the
vector watchdog inhibit allows blind play. I suppose that will be
easier to repair--it's not the C8 rom.
Regards,
Dave
Dave
A couple of corrections:
> I wonder if the vector EPROM access time is just slightly out of
> tolerance?
I meant vector ROM
> I'm measuring -27.4V across C49.
That's actually between the negative side and ground. There's 12V
across the cap.
Regards,
Dave
Matt Hoy
thank you for your contribution.
Dave
No high scores are saved. On powerup, the high-score screeen is
cleared and repainted 3 times, each time with the display
'1. FFFFFF'. Then the machine enters attract mode, with no high
scores. The machine works normally after that, until the next
power cycle/reset.
** Removed the EAROM, same behavior. Replaced with good EAROM, same
behavior. voltage on negative side of C49, -27.4V. Strobes are
present at clock inputs for the 74LS244(M9), 74LS374(L9),
74LS174(N10) and 74LS175 (P9). Then I notice a small solder blob
on pin 5 of the '175 (a control line). Sure enough, the signal at
the shoulder of the pin is different from the output of the data
bus buffer. Desolder the '175, and pin 5 falls off. I don't know
how it got broken in the first place. Replace the '175. Now high
score works perfectly.
** *Now*, the only remaining problem is several regularly spaced
horizontal and vertical 1-bit wide zones, which distort any object
with a slight shimmering effect as it passes through. This is
almost certainly bad X and Y DACs. I'll post if not, otherwise
assume that's correct.
The question is, what the #@%! was the first thing on that board to go
wrong? Probably the DAC's, then the high score, with the botched
repair attempt--perhaps this was a botched high score hack attempt?
Then the audio op-amp. Then the POKEY (just the sounds apparently,
probably not apparent with soft sounds). Then someone plugged it into
the wall socket, vaporizing a trace, frying the diode bridge and
regulators. Perhaps the CPU bought it then. Then the PROM and
crystal were cannibalized. Finally, offered on Ebay as 'untested'
:-).
David.