Re: NEED SOME HELP with a DEAD STAR WARS BOARDSET....

From: Joel Rosenzweig <joel.rosenzweig_at_verizon.net>
Date: Mon Sep 23 2002 - 23:40:28 EDT

Well, this brings back memories of my first Star Wars boardset. It worked
for all of 5 minutes, then failed exactly as you described. What a
maddening experience! I was so excited to have the trench display on my
oscilloscope, just to have it die on me minutes later. I was so ruined by
the experience that I bought 4 more working boardsets so that I'd never be
stuck without a working Star Wars set. :-)

That being said ...

When I have a boardset that is completely dead like you describe, I try to
figure out exactly what the machine _is_ doing, as "completely dead" isn't
always really completely dead.

First things first. After you've checked that you've measured +5VDC at some
IC on the board, you should jump in and take a look for all the clocks that
drive the system. Take a look at the schematics, or look at the test points
that are stenciled on the board. You'll want to use either a frequency
counter or an oscilloscope to measure the frequency of the various clocks.
A scope is better because you'll also get to see if there's any noise on the
clock. Physically check the crystals if the clocks are missing or otherwise
looking poor. I've had several games where the crystal's had rotted away,
leaving just the metal shell sticking to the two stubs of wire! If you
don't have all your clocks, stop here, and fix the clocks. There's no
reason to check for anything else until these signals are looking good.

One you verify the clocks, make sure that the reset signal to the CPU is not
asserted. Consider disabling the watchdog circuit by grounding that test
point. See if the behavior of the system changes. Once you get past the
reset signal/circuitry, put your attention on the address and data bus. You
want to see activity here on (usually) all of the address and data lines.
See if there are any lines that are stuck high or low. Check the outputs of
the buffers and latches that sit on the bus. Make sure that these bus
drivers are actually driving the bus. I've had plenty of faulty 74244s and
74245s. The point is that the CPU runs by performing a fetch, decode,
execute of all the instructions stored in the eprom. If the CPU isn't
fetching code, you're not going to have a lot going on elsewhere. Be aware
that individual IC's sometimes fail, that hang off the bus, that can drive a
data line on the bus high or low and keep it there. One bad device like
this is enough to break the CPU's code fetch cycle because it will corrupt
the data on the bus.

If there's no activity on the bus at this point, replace the CPU.

Perform a continuity test from each pin of each socketed chip to the board.
You've heard that the sockets are terrible, and they really are.
Unfortunately, one bit out of place really can take down the entire machine,
depending on which bit! This is a tedious step, but there's no other way to
tell if you have a connectivity problem.

The ram is the next big item to check/replace. All caveats with respect to
the speed rating apply.

Validate the code in your eproms/proms. Consider replacing them with a
fresh set. I haven't personally encountered this failure mode, but I've
read on vectorlist over the years of people having eproms that read fine in
the programmer but didn't actually work in circuit for a variety of reasons.

Once you have done the above, you have the essential elements that allow the
CPU to function in a bare bones environment. You should be able to run self
test at this point, which if you haven't solved the problem already by doing
the above repair and rework, should help point you in the right direction.

In the future, don't connect your monitor to your untested game board. This
testing activity it best left to an oscilloscope until you know that your
board will come up in some sane state. It's just not worth destroying your
monitor.

If this seems like it will be too big of a hassle, consider just buying
another boardset that's known good. They seem to be on ebay all the time.

May the force be with you.
Joel-

----- Original Message -----
From: "Stempak Arcadius-ASTEMPA1" <ASTEMPA1@motorola.com>
To: <vectorlist@synthcom.com>
Cc: "Stempak Arcadius-ASTEMPA1" <ASTEMPA1@motorola.com>
Sent: Saturday, September 21, 2002 12:46 AM
Subject: RE: VECTOR: NEED SOME HELP with a DEAD STAR WARS BOARDSET....

>
> Joe, (and Others)
>
> Thank you for all the input, appreciate it.
> Will let you all know what happens.
>
>
>
> Arcadius......... :-)
>
>
> -----Original Message-----
> From: Joe Bachmann [mailto:joe.bachmann@newmail.net]
> Sent: Friday, September 20, 2002 10:30 PM
> To: vectorlist@synthcom.com
> Subject: Re: VECTOR: NEED SOME HELP with a DEAD STAR WARS BOARDSET....
>
>
> Grab a copy of the schematics/technical manual here:
>
> http://www.joesarcade.com/schematics/star%20wars/
>
> Stempak Arcadius-ASTEMPA1 wrote:
>
> > Peter, (or anyone else for that matter)
> > *******
> >
> > One other question, since I myself do not have
> > the actual Schematics I was wondering if anyone
> > knows what the Actual EPROM NUMBERS for the EPROMS
> > on the MAIN and AVG Boardset(s) should be as well as what
> > Actual Sockets the EPROMS should go into ???
> >
> > Want to Verify this to ensure Proper EPROMS in the
> > Proper Socket(s).
> >
> > Arcadius......... :-)
> >
> > -----Original Message-----
> > From: Stempak Arcadius-ASTEMPA1
> > Sent: Friday, September 20, 2002 9:58 PM
> > To: 'vectorlist@synthcom.com'
> > Cc: Stempak Arcadius-ASTEMPA1
> > Subject: RE: VECTOR: NEED SOME HELP with a DEAD STAR WARS BOARDSET....
> >
> > Peter,
> > *******
> >
> > Thank You, I really appreciate the Information and will
> > try your suggested actions ASAP.
> >
> > Arcadius.......... :-)
> >
> > -----Original Message-----
> > From: Stempak Arcadius-ASTEMPA1 <ASTEMPA1@motorola.com>
> > To: "'vectorlist@synthcom.com'" <vectorlist@synthcom.com>
> > Date: Fri, 20 Sep 2002 17:02:26 -0500
> > Subject: VECTOR: NEED SOME HELP with a DEAD STAR WARS BOARDSET....
> >
> > >
> > > Hi,
> > > ***
> > >
> > >
> > > How are you all today.....have an
> > >
> > dont you allways ;-)
> >
> > as nobody else has said anything usefull try this,
> > i dont know what test gear you have so try what you can.
> >
> > check the cpu for clock and reset signals,
> > replace the cpu,
> > check the ram,
> > (try moving the ram around,most systems have chip that cant be
> self-tested)
> >
> > get the roms off the net for mame and compare them to your set.
> > (or maybe mowerman can give you the checksums?)
> >
> > replace the address and data buffers around the cpu.
> >
> > IF IT STILL WONT GO - YOUR ON YOUR OWN :-)
> >
> > c.u.
> >
> >
> --------------------------------------------------------------------------
-
> > ** 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.
> >
> --------------------------------------------------------------------------
-
> > ** 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.
>
> --------------------------------------------------------------------------
-
> ** 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.
> --------------------------------------------------------------------------
-
> ** 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.

---------------------------------------------------------------------------
** 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 Mon Sep 23 21:12:26 2002

This archive was generated by hypermail 2.1.8 : Fri Aug 01 2003 - 00:34:09 EDT