[comp.sys.atari.st.tech] Hardware-related csd failure

saj@chinet.chi.il.us (Stephen Jacobs) (09/05/90)

This has the feel of one of those problems that's going to take more brains to
solve than are conveniently assembled in one place.  Here goes:

Doing some development on my MEGA 4 (TOS 1.2, msh, minimal AUTO and accessory
complement).  I went to invoke the Mark Williams csd debugger (gem csd <program>
).  And the screen went blank.  That never happened before.  Some playing
around showed that csd was, in fact, running, and controlling my program.  But
the source code listing, the help messages and the functioon key labels were
either not on the screen or drawn white-on-white.  The screen display of my
program (what is it, F7?, F6? whatever..) was normal.  Once I caught on, I
was able to terminate my program and csd in the usual way.
   Every test I've tried comes out normal.  Other programs run just fine
(although I don't have that many 'challenging' ones handy).  A memory tester
shows the right amount of memory and says it's good.  Comparing csd and msh
to the distribution disk gives a match.  I restored all the system stuff from
a backup just for jollies, and it changed nothing.
   The reason I consider this a hardware problem is that when I switched the
hard disk (ST296 on a BMS-200; configured as 4 15 Meg partitions and a 20
(which last is not used for this program nor for system files); Supra software)
to a 1040 ST, csd worked properly.
   I'm resigned to needing repairs.  But I'd rather get them locally than have
to swap my MEGA with Atari (time and money advantages, both).  The problem 
with that is that this problem seems a bit too subtle for the service guy to
get a grip on.  Anyone out there have a bright idea what I should tell him?

                                     Steve J.