Re: TECH: Williams Stargate Problems

Kirk,

Thanks for the tips. The biggest problem right now is the time to sit down
and start going through everything. I'm 99.999999% sure that it's the CPU
boards and not the Input Widget or ROM board since I tested all the CPU
boards in the same cabinet and some have no input problems at all. As for
the RAM errors, I would like to switch all the ram to 4164 (hope that's the
right number) and get away from the 4116's all together.

I do have a Logic Probe, and I just picked up an Oscilliscope from John
Butler (Thanks John!), so hopefully between those two and my Multi-Meter I
can fix these boards (and not make them worse). This is the fun part for
me. Learning how to do all of this.

Now all I need to do is find a Williams Adapter for my Univid 1000 or make
one... Although I do have a JAMMA harness so I might just build a Williams
to JAMMA adapater.

All I need now is 25 hours in day and I'll be set!

Brien King


"Kirk S." <crayzkirk@xxxxxxxxx> wrote in message
news:O6Sbf.14445$8K4.7170@xxxxxxxxxxxxxxxxxxxxxxxx
> Brien,
>
> From my experience, the PIA on the input widget board and ROM boards cause
> more input errors than anything I've seen. I've also seen some really odd
> displays from bad shift registers (74166 Others/74165 Defender). A bad
> address multiplexor (74153) will cause a really busted up rug and a 1-3-1
> error. Don't forget the address generators (74161/9316) or the blanking
> circuit as culprits. As a last resort on Ram errors, the bank buffers
> (74LS373/374s) can also cause some grief. I had two Bubbles CPU boards
> that
> got random 1-3-1 errors until I replaced all the data buffers. Check the
> clock and crystal area for battery damage, too.
>
> An old HP 10529A logic comparator is really handy on the bank select and
> write enables although a logic probe will show when things are simply
> dead...
>
> In some cases, you can substitute LS/ALS/HCT for the components. Anything
> related to memory addressing or bank control must have the drive capacity
> of
> standard TTL. A 74LS153 burns out immediately as does a 74LS161.
>
> A trick I learned while working on Williams Pinball Driver boards is to
> place the + lead of my multimeter on the ground plane of the board in
> question, set the meter to the diode setting and probe the pins with the -
> lead. You should see pretty normal diode voltage drops on most of the
> logic
> pins. Since TTL typically shorts, look for any really low or really high
> readings. It's helped me quickly locate bad TTL. A logic probe on pin 14
> of the 4116 Rams also will pretty quickly show any dead 4116 Rams on
> partially working boards.
>
> If the board won't boot at all; I usually suspect the CMOS Ram and 74153
> multiplexors. If they short out, they can pull the address lines down and
> prevent the CPU from accessing the game program. I pull the 4116 Ram out
> first so any dead ones can't clog the data buss.
>
> Good luck!
>
> Yeah, Sean has lots of good info on his site...
>
> Kirk S.
> "Brien King" <spammehere@xxxxxxxxxxxxxxxxxxxxx> wrote in message
> news:RIObf.18948$bo.2269@xxxxxxxxxxxxx
>> Would it be possible to have the Diagnostic tool list out possible fixes
> for
>> the problems it finds for the common problems?
>>
>> Such as "Fire Button Stuck ON... check LS74xxx at Location X, Y and Z" or
>> something like that? or would that be more of a waste of time? Anything
> to
>> help speed up repairs :-) I have about 17 CPU boards right now that
>> need
>> fixing. A lot with stuck inputs, RAM errors, and some video problems.
>>
>> BTW: Thanks for all the great info on your site!
>>
>>
>> Brien King
>> http://www.arcaderestoration.com
>>
>>
>> "Sean Riddle" <seanriddle@xxxxxxx> wrote in message
>> news:1131253728.289201.258750@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>> > Steve-
>> >
>> > What's out there now is not yet a diagnostics for Williams boards, it's
>> > just a single ROM that fills memory with different values so you see
>> > colors change on the monitor. It also sets the LEDs to different #s.
>> > Finally, it draws a picture on the screen and waits for you to hit one
>> > of the start buttons. The goal was to have a single ROM for Scott to
>> > test in his ROM board.
>> >
>> > What I'd like this to evolve into is a more generally useful
>> > diagnostics tool: things like RAM test, CPU test, NVRAM test, blitter
>> > test, button test, ROM checksummer, etc. I've been thinking how I can
>> > determine if the CPU board is Defender, Bubbles, Sinistar or other, if
>> > Special Chips 1 or 2 are present, if the watchdog is enabled or
>> > disabled, etc.
>> >
>> > Sean
>> >
>>
>>
>
>


.