Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!ll-xn!ames!amdcad!markg
From: markg@amdcad.AMD.COM (Mark Gorlinsky)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Interrupt driver for parallel printer
Message-ID: <17135@amdcad.AMD.COM>
Date: Mon, 15-Jun-87 12:45:10 EDT
Article-I.D.: amdcad.17135
Posted: Mon Jun 15 12:45:10 1987
Date-Received: Sun, 21-Jun-87 01:54:18 EDT
References: <371@aucs.UUCP>
Reply-To: markg@amdcad.UUCP (Mark Gorlinsky)
Distribution: na
Organization: Advanced Micro Devices, Inc., Sunnyvale, Ca.
Lines: 28

In article <371@aucs.UUCP> paul@aucs.UUCP (Paul Steele) writes:
>I am having difficulty in getting an interrupt driven parallel printer
>driver to work (the way I think it should).  I've done all the necessary
>
>The XT Technical Reference Manual says that an interrupt occurs when
>pin 10 (busy) transitions high to low.  As far as my understanding
>of a parallel printer adapter, this should happen after every character
>is output.  I have written an interrupt driven routine for a QX-10 and
>it worked as I would suspect.
>
>Paul H. Steele      UUCP:      {seismo|watmath|utai|garfield}!dalcs!aucs!Paul

Your problem lies not with you, but with the basic hardware flaws of the 
IBM PC.  It is impossible to generate a true interrupt driven parallel port.
This is because the ACK signal coming back from the printer is tied directly
to the interrupt controller.  Unfortunately the timing of this signal, as
defined for standard centronics interface, is too fast to be acknowledge by
the controller! It is uncanny how one simple flip-flop inserted between the
signal and the controller could have eliminated the problem.  BTW, all clones
out there never bothered to correct the problem, they just copied IBM line
for line.


-- 
 Mark Gorlinsky - AMD Processor Products Division/APPS SQA
 UUCP: {ucbvax,decwrl,allegra}!amdcad!markg or amdcad!phil@decwrl.dec.com
 AT&T: (408) 982-7811
 DISCLAIMER: What's mine is mine, not my employers.