Inverting just one axis on an Asteroids board

From: Doug Jefferys <dougj_at_hwcn.org>
Date: Mon Apr 01 2002 - 14:32:54 EST

On Sun, 31 Mar 2002, Marc Alexander wrote:
> From: "Phil Morris" <phil@morris0.fsnet.co.uk>
> > >Does anyone know if an Atari Asteroids has an undocumented INV Y on it?
> > >I'd like to put Asteroids back in it when I get a board to suit, and
> > >was wondering if it has an INV Y input I can use without modding the board.
> >
> > According to some information in this message:
> >
> http://groups.google.com/groups?q=asteroids+%22inv+y%22+group:rec.games.video.arcade.collecting&hl=en&scoring=d&selm=6cs3q0%2414q%40
> crchh327.us.nortel.com&rnum=6
> >
> > Both X and Y are combined on pin 19 on an Asteroids board. Maybe they could
> > be separated earlier on in the circuit? I don't know. Haven't tried it. No
> > doubt more knowledgeable folks than I can advise. :)
>
> Thanks Phil, I use google all the time, but didn't think of it to do a
> search of newsgroups! I was wondering why Asteroids only had INV X
> listed on spies, now that I see it, INV XY together seems more
> sensible.

First off, props to Gregg for correcting me 4 years ago ;-)

Second, I had a similar issue when converting a Quantum to run on a
horizontal monitor by swapping the X and Y outputs. A hack from Ray
Ghanbari (which I can't locate at the moment) involved a bit of cutting/
jumpering on the board to implement an X-invert (er, Y-invert) capability.

WARNING: Speculation ahead. I have not confirmed this. I'm just
  talking about what I see on the schematics; this is not a confirmed
  hack/fix for the problem.

In Asteroids (schematic, sheet 2b), there's only one set of INV/NINV
outputs, and they're governed by the X/YINV input (pin 19, PCB) and
the transistors at Q16 and Q17.

In order to invert one axis and not the other (e.g. to run Asteroids on a
different-style monitor, or to run it in an inverted cabinet such as
Asteroids Deluxe), it looks like you'll have to figure out which 4016B is
used for which output, and swap the INV/NINV signals on pins 12/13 of that
chip, and make sure that the signals are isolated from each other (so you
don't have transistors driving each other in opposite directions).

For instance, on the schematics I have, I see D12 (X-axis) and B12
(Y-axis) having pin 12 = NINV, pin 13 = INV.

On a board with a 4016 at B12, to invert the Y output for mirror-mounting,
I'd cut traces on the board to isolate B12, pins 12 and 13, and then jump
the NINV trace (that used to go to B12-pin12) to B12-pin13, and similarly
jumper the INV trace (formerly connected to B12-pin13) to B12-pin12.

I'd then make sure I hadn't inadvertently "cut off" the INV and NINV
inputs to the other amp. Not having the board in front of me, I can
think of three ways in which the traces could be laid out:

If the board is laid out:
   B12 <----> INV/NINV <----> D12 (e.g. a "Y" connection to B12 and D12.)
   - Then you're done. Probably only had to cut one trace on each B12 pin.

If the board is laid out:
   INV/NINV <----> D12 <-> B12
   - Then you're done. Probably only had to cut one trace on each B12 pin.

If the board is laid out:
   INV/NINV <----> B12 <-> D12
   - Then you'll need to reconnect the original INV/NINV pins to D12,
     or something weird will happen because D12's "invert" signals are
     floating in mid-air.

(Note that all three wiring/routing possibilities are reasonable; the
 Atari folks had no reason to believe you'd want to invert one axis
 and not the other. If NINV/INV are assumed to be the same on both
 B12 and D12, why not route the board with two long traces and two short
 traces, rather than four long traces?)

Disclaimer 1: I'm working from a PDF of a photocopy of the schematic
and don't have the board near me. Chip locations and pin assignments
may vary depending on the revision of the board.

Disclaimer 2: There may be an easier way to do this, namely fiddling
around with the TL082 that does the final output.

Disclaimer 3: I haven't had enough coffee this morning. This post is
a jumping-off-point for discussion, *not* something I'd trust without
verifying :)

Later,
Doug.

-- 
 dougj   |
   @     |
hwcn.org |
---------------------------------------------------------------------------
** To UNSUBSCRIBE from vectorlist, send a message with "UNSUBSCRIBE" in the
** message body to vectorlist-request@synthcom.com. Please direct other
** questions, comments, or problems to neil@synthcom.com.
Received on Mon Apr 1 11:37:36 2002

This archive was generated by hypermail 2.1.8 : Fri Aug 01 2003 - 00:33:58 EDT