*BSD News Article 16669


Return to BSD News archive

Newsgroups: comp.os.386bsd.questions
Path: sserve!newshost.anu.edu.au!munnari.oz.au!uunet!math.fu-berlin.de!easix!knobel!andreas
From: andreas@knobel.GUN.de (Andreas Klemm)
Subject: Re: Elusive SCSI problems (Adaptec 1542C) <snarl!>
Organization: Andreas Klemm, D-4040 Neuss 21
References: <1993May14.004015.23497@sophia.smith.edu> <hastyC708F2.CGH@netcom.com> <C70nBJ.387@world.std.com> <1993May14.155303.6702@sophia.smith.edu> <1993May28.153410.10495@cs.wisc.edu>
Message-ID: <1993May31.091928.822@knobel.GUN.de>
Keywords: Adaptec 1542C, Quantum, SCSI, 386bsd
Date: Mon, 31 May 93 09:19:28 GMT
Lines: 87

In <1993May28.153410.10495@cs.wisc.edu> jcargill@oka.cs.wisc.edu (Jon Cargille) writes:

|In article <1993May14.155303.6702@sophia.smith.edu> jfieber@sophia.smith.edu (J Fieber) writes:
|>
|>When I called adaptec tech support, the person *kept* asking if I
|>was using external devices.  I then saw this post and understood why.
|>
|>I don't have anything external, yet I still have problems.  
|>
|>  1542c ----- Q P105s ----- Q PD210s ----- Archive Viper
|>  ID#7          ID#0          ID#1             ID#2
|>
|>the 1542c and the archive viper have termination on.  Turning it
|>off at either end has no effect, turning it off it both ends
|>makes nothing work.  (surprise.)

Could it be possible that your internal scsi cable isn't ok ?
I use twisted cable. They should be better.
Or another idea: Try connecting only 1 or two devices with very
short cable and proper termination.

	|-1542c ----- Q P105s -|
	|-1542c ----- Q P105s ----- Q PD210s -|

Something I read from net: maybe it helps......

+----------------------------------------------------
| From comp.os.linux Mon May 31 11:06:43 1993
| From: solca1@urz.unibas.ch
| Newsgroups: comp.os.linux
| Subject: Adaptec 1542C Giga HDD
| Message-ID: <1993May25.154059.42934@urz.unibas.ch>
| Date: 25 May 93 15:40:59 MET
| 
| Hi there ,
| 
| I have a 486DX2-66 with an Adaptec 1542C . The Hard disk attached is a
| IBM 663E12 1Gb . The controoler is SCSI-II like the HDD (f-Scsi-II)
| The Problem is that Linux doesn't sense my controller which means that I 
| cannot boot Linux from the disk purchased from Yggdrasil.
| 
| The controller I haved earlier Adaptex 1542B worked fine. The HDD was a 
| conner 200Mb . What is the Problem . 
| If someone have a Idea Please contact me .
| 
| Thanks in advance 
| Jan.
+----------------------------------------------------

+----------------------------------------------------
|From comp.periphs.scsi Mon May 31 11:09:35 1993
|Newsgroups: comp.periphs.scsi,comp.sys.ibm.pc.hardware
|From: leop@percy.rain.com (Leo)
|Subject: Re: AHA-1542C Bug on external connector???!!!???
|Date: Sat, 29 May 1993 06:07:35 GMT
|
|>In article <1993May15.092748.19463@arbi.Informatik.Uni-Oldenburg.DE> Andreas.Lemke@arbi.informatik.uni-oldenburg.de (Andreas Lemke) writes:
|>>In c't June '93 (German mag) I read about a bug on the 1542C which
|>>concerns direct I/O programming of the adapter for devices attached
|>>to the ext. connector which yields to writing errors on large data
|>>blocks. The ASPI driver (for DOS) supplied by adaptec contains a workaround
|>>but third party software (tape backup programms *AND MAYBE LINUX*)
|>>do not contain this bug fix.
|
|I sent the article to Adaptec and they contacted the magazine.  It
|seems that SCSI-2, and therefore the 1542C, is VERY fussy about the
|external cable used.  Here are some excerpts from mail I got from the
|editor of the magazine yesterday:
|
|(about Adaptec claim that the problem must be the cable used)
|>The man is right! After a lot of messurement, we found out, that the
|>AHA-1542C uses very fast rising signals on the REQ-Line witch requires
|>kables that meet the SCSI-2 specification even when you use SCSI-1
|>drives.
|
|(about why it worked with ASPI under DOS, but not UNIX)
|>Well, it make sense. Cause ASPI-based Software often uses smal
|>block-sizes (below 32 KByte) and the error only occure, if you use
|>large blocks. In our next issu (c't 7/93) we will publish an article,
|>witch describe exactly, whats going on with the C-type and how you can
|>avoid problems with external equipment.
+----------------------------------------------------
-- 
----  Andreas Klemm  -------    /////// andreas@knobel.GUN.de \\\\\\\
private : +49 2137 12609        D-4040 Neuss 21 (Norf), Germany
at work : +49 2173 3964 161     Wiechers & Partner   Datentechnik GmbH
Telefax : +49 2173 3964 222     Abteilung Unix Support, D-4019 Monheim