-----Original Message-----
From: Bruce Boschek <[log in to unmask]>


>In <[log in to unmask]>, on 03/15/99
>   at 06:51 AM, Michael Beechey <[log in to unmask]> said:
>
>>Dear Friends,
>
>>I'm putting together a small network, for classroom use, no essential
>>data transfer, maximum distance to hub 5 meters ( about 15 feet), using
>>cheap NE2000 clone cards, 15 machine max.
>
>Michael,
>
>Coax and BNC connectors seem to win hands down in this kind of a
>situation.

>Twisted pair can be made to allow higher data-transfer rates, but if
>understand your requirments this will be of no notable benefit to you.
>On top of it you will need much more "equipment," a wall socket for each
>machine, transciever cables, a hub and diverse connectors and crimping
>tools, plus a bit of testing equipment if you don't want to have lots of
>headaches.
>


I totally agree that coax would be ideal for his situation but I think
that you have gone way overboard in your explanation of what is needed for
a -T network.

I recently supervised the installation of a 380 node, multi-building,
campus LAN.  I specified pre-terminated custom length fiber cables between
the concentrator and the remote wiring closets (none of which Mike would
need), but all 380 nodes were connected with one tool -- an RJ45  crimper.

Wall sockets are a nice "window dressing" but are not required and, IMHO,
reduce the reliability of the network (each connector/connection adds one
more possible failure point into the net)

I have an expensive TDR scanner for cable certification, but (in most
cases) a us$50 multi-pair tester and a set of walkie-talkies takes care of
all but the most elusive problems (assuming you use quality cable and
follow good installation procedures).

            Do you want to signoff PCBUILD or just change to
                    Digest mode - visit our web site:
                    http://nospin.com/pc/pcbuild.html