🚀 go-pugleaf

In article <3kkDAHACULwwEwNj@davita.demon.co.uk>
           David@davita.demon.co.uk "David Reid" writes:

>I notice that the Etherlink card has a D connector (15 way?) as well as 
>the BNC connector. Is it an alternative (better?) way of connecting the 
>two cards?

No. These are for an alternative flavour of ethernet connection ( AUI ? )
which runs at the same speed but uses more expensive older technology
cabling.


> Assuming that I use the BNC connectors, why are the T-pieces 
>and 50ohm loads necessary? Why not just put the end of the cable
>straight onto the coax?

Here we go again. This may be a little hard to absorb...

  The cable is a transmission line. In fact all cables are transmission
  lines but this is only an important design issue for high speed signals.

  Signals are reflected from points in a transmission line where there are
  changes of characteristic impedence. Thin wire ethernet cable has a
  specified characteristic impedence of 50 ohms.

  There is no more dramatic change characteristic impedence than an
  unterminated wire end, so to avoid the ehternet signal being reflected
  back off the end of the cable, it is run into a 50 ohm terminating
  resistance.

  Now, you could build these terminating resistances into the card but
  then you could only ever link two nodes as you only want to terminate
  the _ends_ of the cable.

  In fact, if you build the terminating resistances into the card
  then you can only link two nodes on one piece of wire but you
  canget round that by using hubs. You'll see this sort of solution used
  with screened and unscreened twisted pair ethernet cabling using RJ-45
  plugs and sockets.

  
>As far as I can tell it *should* be easy to setup using coax, just plug 
>it all together, tell windows about the card and Robert is ones parents 
>sibling as they say. I'm mainly just curious.

Yup. Go for it !

-- 
John Beardmore