[comp.sys.apple] Problems in 80 columns

kamath@reed.UUCP (Sean Kamath) (07/04/87)

Hello again, world.

Whilst trying to write the DCOM drivers for my Cat, I ran into a bit of
a problem.  When I left the //e on for about 4-6 hours, I started
getting junk on the 80 column screen.  By 10 hours, things were in
seriously bad shape. In an old back issue of Open-Apple, I remembered
reading about a problem chip on the motherbaord.  IT's at location B2,
and it's a 74LS245 tri-state bus buffer chip.  It seems that Apple used
cheap singapore parts in their //e's, and this chip flips out when it
get's hot.  I read that they used a Texas Instruments chip on the 80
column card, and that the problem could be alleviated by swapping.
Alas, my micromax 80e uses an even cheaper chip.  So I switched the 245
on my other //e's 80 column card, and the problem almost went away.  I
went out and bought a 245 (all of these are TI chips, exept the micromax
80e, which was nobrand) and then it reall did all go away.  However,
when I shut my 'e off for the night, the next day it wouldn't boot my
assembler. switching chips didn't seem to help, so I went all out and
get the diagnostics card out, the diagnostics disk out, and let 'em
loose.  Nothing wrong.  Worse yet, after about 15 minutes of being on,
there were no problems, or almost none.  Luckily, Apple supplies their
dealers with a TI 245 chip.  I swapped out the one from the 80 column
card, and I haven't had a problem yet.

BTW, I have a fully loaded //e, with all but one slot (7) filled, and a
system saver.

Moral of the story, if you are having problems, then get a couple of 245
made by TI, and put em in there.  If that doesn't solve the problem,
*then* try the dealer.

Also, I hear that a few //e's were manufactured with 74LS109's at
location C1.  It can be a 74F109 or a 74S109, but not an LS109.  they
don't work at all.

Open-Apple had a way to test for this problem (with the 74LS245).  It
consisted of going into appleworks and created and deleting files all
day.  I did it like this:

10 PR#3
20 FOR I= 1 TO 80 : PRINT "x"; : NEXT : PRINT SPC(80): GOTO 20

and let it run.  It only to my //e about 2 lines to bomb out when it was
warm.  I didn't have any problems after I switched the chips.  I suggest
you let this run for at least 10 hours to be sure.

Sean Kamath
tektronix!reed!kamath