Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!mips!spool.mu.edu!munnari.oz.au!yoyo.aarnet.edu.au!sirius.ucs.adelaide.edu.au!levels!etac
From: etac@levels.sait.edu.au
Newsgroups: comp.sys.amiga.hardware
Subject: Re: Tape Drives, Parnet & CIA's
Message-ID: <16150.2801fcab@levels.sait.edu.au>
Date: 9 Apr 91 08:10:59 GMT
References: <16081.27f751e7@levels.sait.edu.au> <1991Apr3.193247.13318@jato.jpl.nasa.gov> <4290@bnr-rsc.UUCP>
Organization: University of South Australia
Lines: 59

In article <4290@bnr-rsc.UUCP>, mascot@bnr.ca (Scott Mason) writes:
> In article <1991Apr3.193247.13318@jato.jpl.nasa.gov> jdickson@jato.Jpl.Nasa.Gov (Jeff Dickson) writes:
>>In article <16081.27f751e7@levels.sait.edu.au> etac@levels.sait.edu.au writes:
>>>
>>>        My flat mate also has an A1000 and we're trying to run
>>>Parnet to network the two machine.  We've made two cables
>>>one about 10 inches the other about 15 metres. The short one
>>>works fine but the long one just won't go. I've checked the
>>>cable a thousand times. I can't see whats wrong.
>>
>>      10 inches VS 15 metres (approx 1 yard X 15 = 45 feet). Try attenuation.
>>
>
> I would recommend a visit to your local electronics supply or hi-end audio
> store. Select the cable with the lowest capacitance per metre (that you can
> afford). Each time the output switches, the driver must charge the cable
> to the new logic state. The capacitance of the cable will have a direct
> affect on this switching time. If you have routed near other electrical
> appliances, such as power supplies or monitors, the cable may be picking
> up considerable noise. A shielded cable will help eliminate these types
> of problems. The shield should be grounded at one end *only*, and left
> unconnected at the other end. This prevents currents from flowing in the
> shield. If you can control the data rate, a lower data rate should be able
> to communicate over a longer distance.
>
>>>        Is this all due to just the poor drive of the 8520 CIA
>>>chips or what?
>>
>
> 15 metres is asking a lot.

Thanks guys for your comments.  Sounds to me like there are two basic
problems.
         1)  The 8520 have bugger all drive, and were probably neverl
               designed to drive cables directly. (buffering was obviously
               assumed to be used).
             15 metres is not really a lot. That should be a piece of
               cake for the serial port.  (I have tried it,it works
               fine, and I know the port is buffered.)

         2)  PARNET's error checking is obviously not good enough to
               let you know the connection is not there or is
               unreliable. (This is not a flame, just an observation)
>>>
>>>Andrew Chalmers.
>>>
>>>Digital Communication Group D.C.G.
>
> You're kidding, right :)

                I jest you not!

>
>>>University of South Australia
>
>
> --
Thanks
        Andrew Chamlers.