Re: TECH: Help with collection of dead Tempest PCBs

From: Jess Askey <jess_at_askey.org>
Date: Mon Jul 16 2007 - 01:31:05 EDT

Here is my basic approach to fixing CPU/VG issues on Atari vectors... I
will generalize and you can ask more specifics if you want... this
process works pretty well for me...

Firstly, if you have a 100% working boardset, you can use that to
determine if the CPU or Mathbox PCB is bad on your non-working boards. I
have never fried a working boardset with another non-working mate. Im
sure others have but it is pretty damn rare. You might as well use that
approach as the first step in a divide and conquer approach.

1. Resolder the interboard molex connector, suck off the old solder and
get it fresh. You might want to do this on your 100% working boardset
too because after yanking that connector off and on a few times, this
could make your nice 100% board stop working too.

2. Power Supply... make sure the +5V is solid on both boards and you
don't have an issue with burned connectors causing the issue. When
testing on the bench, use a known good switching power supply. You can
use the +12V for the 10.3 unregulated. You don't need the +/-22V for
testing the CPU and Vector logic.

3. Can you run self test? Maybe it will just tell you what the problem is.

4. Check the clocks.... 50% of the Atari boards I fix are due to the
crystals being broken since they stand up vertically, this was lovely
when the PCB was safe in a cabinet but doesn't bode well when the PCB's
are in a stack. I just make sure that the CPU clock is pulsing with a
logic probe.

5. Check the reset line... it should pulse low and then stay high. If
you dont' get a nice low blip on the logic probe, the CPUwon't start
properly. If it doesn't act right, then check the power up circuitry
which is related to the 10.3 unregulated voltage.

6. Start replacing all socketed chips from a known good board. Be
careful when inserting chips as if you bend a pin, you will start
causing more problems and destroying your nice spare working PCB. :-(

7. Probe the Address lines and Data lines, they should all be pulsing.
If not and you have already replaced all the socketed chips, then
suspect the LS245's and the RAM. Be careful pulling IC's out, if you
start mangling traces, then again, you are moving backwards quickly.
When in doubt, protect the PCB and sacrifice the IC since RAM and TTL
are cheap. You can cut the pins off individually and then remove them
one by one if you have a nice pair of tiny wire cutters.

8. Is the Reset line pulsing? If so, you have a barking watchdog...
there is something wrong with the CPU that the programmers didn't plan
on happening. You problem is most likely in the CPU section for Tempest.

Those are my standard starting steps. From there, it gets a bit varied
depending on what you find. :-)

good luck

jess

Doug Hill wrote:
> Hello gang.
>
> I have a tangential question for the vector list but I'm hoping you'll humor me. In restoring two Tempest games over the last year I now have a collection of Tempest PCBs that basically don't work. With the help of this list [and all the past posts] I now have two solid WG6100s to which I can attach these dead boards but I am at a loss as to where to start with the repairs on the game boards.
>
> When hooked up into one of the working cabinets most are just dead [no flashing start buttons, test switch does nothing, and obviously spot killer is on]. As far as I can tell the boards look complete [when compared to my two working board sets] and I have done the usual examination of the interconnect cable and the connectors on both PCBs.
>
> I've googled around quite a bit and almost all of the information online assumes the game PCBs are working and the WG6100 is the culprit - I now have the opposite situation. Is there any recommended reading as to where to start with these boards or do any of you guys have some places I should start?
>
> Thanks for any tips you may have and again my apologies for the slightly off-topic question...
>
> -doug
>
>
>
>
> ---------------------------------------------------------------------------
> ** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
> ** Please direct other questions, comments, or problems to chris@westnet.com
>

---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Mon Jul 16 01:30:55 2007

This archive was generated by hypermail 2.1.8 : Mon Jul 16 2007 - 09:50:00 EDT