Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!lll-winken!uunet!mcsun!cernvax!chx400!chx400!urz.unibas.ch!infocenter
From: infocenter@urz.unibas.ch
Newsgroups: comp.protocols.appletalk
Subject: LocalTalk Cards on 33 MHz 386
Message-ID: <1991Mar7.130114.1427@urz.unibas.ch>
Date: 7 Mar 91 12:01:14 GMT
References: <17320@crdgw1.crd.ge.com>
Organization: University of Basel, Switzerland
Lines: 45

In article <17320@crdgw1.crd.ge.com>, news@crdgw1.crd.ge.com (USENET News 
System) writes:

> Compaq deskpro.  The software behaves as if the transceiver is not attached
> to the (card) which it is,  and all the hardware works on other systems.  I
> called Sitka (or whatever they call themselves) and was told that this
> was a known problem and the only solution is to by another card which is
> under development.  I can construct a number of scenerios underwhich
> a really fast CPU would cause problems,  but the Flashtalk card is the
> first device that has failed to work in this system.  Has anyone else
> experienced this problem? Better yet has anyone found a solution.  I
> quickly scanned a dissasembly of atalk.sys looking for problem areas.  I
> noticed a couple of things that looked suspicious,  but I thought that I would
> inquire of the network before I tried patching someone else's device
> driver.  I'm not even sure that the problem lies in the driver.  Sitka
> tech support was neither very technical nor very helpful.
> From: dixon@sagittarius.crd.ge.com (walt dixon)
> 
> Walt Dixon		{internet:	dixon@crd.ge.com	}

I am fighting with a similar problem. We have original LocalTalk PC Cards 
running in different machines (286/6 ..25, 386SX/16, 386/25, etc.). 
But those suckers won't do it in my 386/33 machines ( all 386/33 have the 
same board, we tried changing BIOS, this didn't help). 
We tried all permutations of possible interrupts and addresses (220h & 240h) 
... they won't do it!

So I tried DaynaStar LT200 PC LocalTalk Interface board, running with the 
original Apple AppleShare PC Software (vers. 2.0.1).
This works just fine!
LT200 has a nice feature: with jumpers you can choose a lot of different 
addresses. Well, I guessed that this was the problem with Apple's LT-Card.
So I set the Dayna on 220h and on 240h. Both worked with no problems !?!?!

Bus speed?
My 33s run on 8 Mhz, exactly the same as my 25s etc. (I do not believe data 
sheets, I MEASURED it !!!).
The only difference: the 33s' bus signal show more overshooting and 
undershooting than those of the other machines. It really oscillates.
Well, this could explain that the card gets triggered twice, but since all the 
other stuff on the bus works fine ... I do not believe this is the problem.

So,  W H E R E  is the problem ?

...............................................................  Didi