Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site mhuxm.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxl!mhuxm!lmb
From: lmb@mhuxm.UUCP (breda)
Newsgroups: net.micro.pc
Subject: Odd bug with A/D on PC - help!
Message-ID: <1226@mhuxm.UUCP>
Date: Tue, 17-Apr-84 15:51:51 EST
Article-I.D.: mhuxm.1226
Posted: Tue Apr 17 15:51:51 1984
Date-Received: Wed, 18-Apr-84 08:09:23 EST
Organization: AT&T Bell Laboratories, Murray Hill
Lines: 23

I have an intermittent problem with a Tecmar
Labmaster A/D board in a Compaq PC compatible.
The hardware test software is written in
assembly language and is called by a C program.
It simply reads the A/D, fills a 1000 element
stack and then returns.  Everything works fine when
the Compaq is first turned on.  (The test signal
is a sinusoid.)
However, after a few runs, the high byte
is read as a low number, when it should be
0xF.  This only happens with low negative
numbers.  The outputs from the A/D are
okay, but they are not getting to memory.
Particulars:
	- 512K bytes on-board memory
	- CI-86 C
	- also have Tecmar Winchester 33/share expansion unit
Any suggestions?  Anyone else run into this?
Thanks.

Loreen Breda
AT&T Bell Laboratories	(201)-582-7009
mhuxm!lmb