Re: Battlezone Half Screen, please help if you can

Ok, the /ZBLANK signal in the AVG is one of this clever BZ/Atari tricks
which is designed to make the BZ look better than it is. What it does is
to check any vector (x or Y) against a set threshold (or clipping level)
and if the threshold is exceeded, then the vector is blanked at that
point. Suppose you have a pyramid block and you drive up to it, then as
the vectors ramp up to their full extent, they would be drawn in the
instrument panel display, in fact if you force /ZBLANK high, this is what
you will see on the display and in fact driving up to a block is part of
the setup delivery check which is listed in the operators manual.

It works like this: The logic calculates the endpoint of the vectors, but
cannot control (or probably have time to do) clip the vectors at their
peaks. So, as the voltage of either X or Y ramps up towards the threshold
the vector is visible, but after a limit, the vector still draws, but it
is blanked so it is hidden in the intrument panel. At the start of any
game screen refresh, the vector code sets the treshold by executing a
special vector instruction. Normally, for BZ it says, set the +Y to say
about 60/70% of the vector range (for a +/- 10 volt range this means set
to about +6 volts), set clipping on the -Y to -10v, +X to +10 V and -X to
-10 v. (Note that this is for a game screen, for a self-test, clipping is
set to maximum - or should be). It does this by setting an output on the
vector DACs corresponding to the vector clipping level and thowing the
analog switches in turn (signals /HST & /LST) so that the caps C94/C69 -
C93/C70 hold this voltage set. The switches are then set off, but the caps
hold their value, as the following voltage buffer op-amps are very high
impedance inputs. The threshold levels are applied to one input of the
comparators, B12 (D12) so that should the vector being generated exceed
the threshold, then /WNDBLK is set low. effectively all comparator outputs
are wired-or together, but in practice only one output will be low. This
then hides the vector at the point of clipping.

Setting the clipping level is the first operation usually at the start of
a vector run. You can see 4 large deflections (+y, +x, -y, -x) at the
start of any sequence.

What you need to do is to check out the section for your half screen, you
might need to be a little creative. 1. you don't need the diodes, they are
there for protection, you could lift them out, 2. you can check the
switch logic drive, ie, /HST /LST signals, they will only be low for a
very short time at the start of the vector cycle, 3. check the voltages on
the voltage follower outputs (C11) will, they will be as described above
when the game is running, or +/-10 volts in selft-test mode. You won;t be
able to check the caps, as you will probably discharge them. Finally, if
still no progress check every passive component around the amplifiers and
look for board shorts/baseplane shorts.

Persevere, you can do it!!!

Hope this helps...

+Tim+

PS: Forcing signals high/low like /ZBLANK is only for diagnostics, you can
learn a lot from this just on a temporary basis until you figure out whats
really going on. From your notes /WNDBLNK is the key !!!

MKDUD@aol.com
28/02/03 05:22

 
        To: ygroups@andora.org.uk
        cc:
        Subject: Re: VECTOR: Battlezone Half Screen, please help if you can

Hello Tim,

     Thanks a bunch for the response. Here is what I have so far:

>>Can we assume that theAVG - XY outputs look fully bipolar on a scope,
before they get to the monitor?<<

Yes, I am seeing the full picture on the scope. Problem is certainly in
the Z section.

>>If you are sure it is the Z-output have you looked at the window
clipping
circuit? you say that p12 of B7 is HLP (is that high-low pulse?)<<

Yes, p12 of B7 (LS86) is high-low pulse.

>>and you say it should be both be high, well in this case, the clipping
circuit is
clipping when it should not, so suspect that either H10 faulty, or analog
switch D11 faulty.<<

Replaced both H10 (LS74) and D11 (DG201, cross to LF13201). Same problem.
Also replaced C11 (TL084), D12 (LM319), and B12 (LM319). No effect.

>>If you have socketted B7, just force /ZBLANK high and
see what you get.<<

When I force /ZBLANK high, I get a normal picture. This seems to fix the
problem. HOWEVER, what does /ZBLANK do, and when would it have to go low?
I really don't want to call forcing /ZBLANK high a fix.

Also, P12 of B7 (LS86) is HLP, and I can't figure out why. I checked all
diodes (CR3-CR6, CR7-CR10) and they are all good.

I also replaced P9 (LS164) and N9 (LS02). Still no luck.

I really don't know what else to do. My next plan is to try to use the
scope to find out why /WNDWBLNK is HLP by tracing back through B12 & D12
(LM319's), C11(TL084), and D11 (DG201).

Any further help would be appreciated.

Thank You,
Michael Kelley

---------------------------------------------------------------------------
** 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 vectorlist-owner@synthcom.com.
Received on Fri Feb 28 04:08:27 2003