golfrock wrote:
> Thanks for the comments.  The machine is fine with anything else.
> Warm or cold.  So I don't think the mobo RAM is flakey.

That may or may not be the case.

RAM failures, including intermittent or temperature-dependent
failures, may be pattern sensitive, and, in any case, require that
the system or a program actually use the flaky RAM.

It is not at all uncommon for bad RAM to show up only under specific
conditions of use.  A RAM test is designed to exercise *every* address
using several data and address patterns that tend to expose most RAM
problems.  It can even help if you suspect temperature-dependency to
use a hair dryer to heat the parts and some cooling spray to cool
them down (to confirm that the error comes and goes).

The most fundamental troubleshooting technique is called "moving the
problem", and involves changing parts in a combinatorial way so that
the bad "chunk" can be identified.  (If there is more than one bad
chunk, this strategy becomes much more difficult.)

So, to eliminate the disk as a source of the problem, you could try
running it on another (essentially identical) machine and see if the
failure moves with the disk.

To eliminate RAM in the same way would require moving the RAM of the
failing machine to another machine to see if that makes it fail in the
same way.  Or the equivalent "negative" approach would be to replace
the suspect RAM with known-good RAM and verify that the problem does
not occur (though this approach does not actually *confirm* that the
original RAM was bad, it is a common technique).

In the latter case, a confirmation test would be to replace the original
RAM and re-confirm the presence of the error.  If the error no longer
occurs, then it may be that the act of removing/replacing the RAM
actually removed the fault (like a bad solder joint).

The problem is that your RAM is soldered to the board, and so is not
easily replaced.

That's why a thorough RAM test (more thorough than the built-in self-
test) is the most practical way to proceed to determine if the RAM is
bad.  If it fails the test, then it's bad.  Unfortunately, if it passes
the test, it's not necessarily good--maybe the test just didn't find
the error!

If you are faced with a situation where any one of several things may
be faulty, you need to consider both the likelihood of each thing
failing and the difficulty of testing for its fault.  Then proceed
by doing the tests in the order of decreasing probability of success
and increasing difficulty to minimize time and effort in finding the
fault.

"Hunches", unless based on solid experience, are usually just time-
wasters.

> I have two original Joust disks and each one does that.  I thought it
> was a sector going bad when I found the second disk and it worked
> properly the first time.  Then later, it did the same thing.

So it seems unlikely that the problem is the disk--just as unlikely
as two disks developing identical problems.  ;-)

Of course, the disk drive is common to both situations, and it may
have a problem that is only detected by Joust...  Or the computer,
or any of its parts, like the RAM.

> I couldn't play these at all on a ROM 3.  That's why I setup a ROM 1
> for these older games.  Does anyone know of a KVFJ switch?  That would
> be kayboard, video, floppy, joystick.

You're kidding, right?  ;-)

-michael

******** Note new website URL ********

NadaNet and AppleCrate II for Apple II parallel computing!
Home page:  http://home.comcast.net/~mjmahon/

"The wastebasket is our most important design
tool--and it's seriously underused."
--- SBBSecho 2.12-Win32