From: utzoo!decvax!harpo!floyd!cmcl2!philabs!sdcsvax!sdccsu3!sdcvax!bmcg!rasp
Newsgroups: net.micro.atari
Title: Re: CRASH!
Article-I.D.: bmcg.209
Posted: Wed Feb  9 18:41:35 1983
Received: Sat Feb 12 06:57:08 1983
References: utzoo.2788 uwvax.743



Here is some accumulated lore about crashes and glitches on the
ATARI:

	1. When you clean the contacts, clean all of the contacts!
	This includes CAREFULLY removing the ROMs from their
	sockets, GENTLY erasing oxidation from the pins, and
	reseating them.

	2. When you use an eraser to clean off the contacts, be
	VERY gentle. The precious-metal coating is extremely thin;
	and, when that gets rubbed off you're in for a lot of
	oxidation problems.

	3. As a rule of thumb, with gold-to-gold conections, the
	problem won't be oxidation... just accumulated dirt. The
	clean-up period for this should be on the order of months.
	For gold to silver there is a mild oxidation problem that
	could require clean-up on a one or two month basis. If it
	is gold-or-silver to something else, like German silver or
	solder, the service interval may be days or weeks depending
	upon ambient temperature and atmospheric contaminants; e.g.,
	if you live near a paper mill or a rubber plant with the
	resulting corrosive atmosphere you have got a real problem.

	4. One other bit of unpleasantness is that some alternate
	sources for add-in memories had a problem EXECUTING code
	out of high memory. It turns out that the 6502 makes
	adjacent accesses to memory when fetching code while it
	makes less frequent access to memory on data fetches.
	This means that a memory test that only accesses data in
	memory won't detect the problem. What is needed is a test
	that moves a piece of relocatable code through the area
	under test and executes it. The only problem there is
	that it won't pick up pattern sensitivities. Oh well!
	The nice part about it is that the manufacturers own up to
	the problem when asked, and their warranties are still in
	effect.

	5. Still another little mystery is the misbehavior of the
	system if you fool around with the joystick at inopportune
	times; e.g. during disk accesses, etc. I'm not entirely
	pleased with the explanation; but, it could be an electro-
	magnetic interference problem. Solutions are to avoid
	keeping your computer in a room with nylon carpets (static
	build-up) or any other insulating material. Of course, as
	far as diddling with the joystick while accessing the disk
	goes, it's simply a case of: "Doctor, it hurts when I do
	this."  ... " So, don't do this!"

Good luck with your machines.

				Ron Perloff
				Burroughs
				Micro Components Group
				PO Box 28810
				San Diego, CA  92127
				(619) 451-4519