Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!shadooby!samsung!usc!ucsd!ucbvax!CCIW.BITNET!U009
From: U009@CCIW.BITNET
Newsgroups: comp.sys.atari.st
Subject: Floppy Problems After 4 MB Memory Upgrade on OLD 1040ST
Message-ID: <8911180803.AA17628@ucbvax.Berkeley.EDU>
Date: 17 Nov 89 16:55:16 GMT
Sender: usenet@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 52

Received: from CCIW.BITNET by McMaster.CA; Fri, 17 Nov 89 11:45 EST
Date: Fri 17 11 89 11:10:00 EST
From: U009@CCIW.BITNET
Subject: Floppy Problems After 4 MB Memory Upgrade on OLD 1040ST
To: Info-Atari16@Score.Stanford.edu                                 (

I posted a request for info to the net a while ago and I thought it
might be time to update y'all:

> Back in early May, I purchased and installed a Datafree Industries 4
> MB upgrade card into my (old) 1040 ST.
> .....
> Since I use the system about 95% in the Mac emulation mode with a
> Spectre cartridge, I didn't notice anything wrong at all. I could read
> and write floppies (Spectre and Mac format with a Translator-1)
> without problems. It appears, however, that there is a problem on the
> Atari side under GEM. If I try to copy files to the floppy, the system
> will most often re-boot (like I hit the RESET switch) at some point
> during the first file.
>
> .... info about power supply deleted ....

and from some email I sent to glk3593@iuvax.cs.indiana.edu

> .....  It might be that my
> TOS (quite early) doesn't support >1MB properly. I am expecting the
> V1.4 TOS ROMs soon and if it makes a difference, I will post to the net
> and probably directly to you.

Well, the 1.4 ROMs arrived and after installation, the system acted
exactly the same. Spectre operations worked flawlessly while desktop
copies under GEM bombed quite reliably.

I recently opened the 1040 again and removed the high 2 MB bank and
re-enabled the old .5 MB bank on the motherboard. The problems have
apparently completely disappeared!!!! I copied >80 small files to a
floppy (till it errored out because it was full!) several times under
the desktop on GEM and the Spectre operation hasn't given any problems
either (except for the reduced memory size).

Might I ask if Mr. Pratt and/or Mr Townsend might enquire of the
'hardy boys' (thanks Tracy) in the back room if there was an early
problem with the MMU which might have shown up at exactly 4 MB of
memory (an off-by-one error in the decoding or something)?

I'm thinking of re-cycling the 1040 (sans extra memory) thru the
classifieds and looking for a current model if there isn't a simple
chip swap fix.

Regards, Stu Beal, VE3MWM, (U009@CCIW.BITNET),
National Water Research Institute,
Burlington, Ontario, Canada.