[plug] home cat5 issues

[Brad Campbell]

Mark J Gaynor wrote:
> I have not heard of a modified star topology, a star is a star. I would say
> that you
> have seriously changed the characteristics of the cable by placing sockets
> along
> its length. If the sockets were a normalized type (not come across cat5
> normalized
> sockets either) then I would say the idea would work. A normalized socket
> breaks
> the connection at the point of connection. I would say that stray capacity
> is your
> worse enemy here, resulting in a sad lack of speed.

Not so much stray capacitance, but reflections from un-terminated cable spurs.
Sorry to be the bearer of bad news, but what you want to do _will_not_work_reliably_ever_.

Cat 5/5e/6..and so on is strictly point to point. The length limitations are to do with transmission 
speed and reflections, and you will find that if you go from one end to the other it will work, but 
any point in the middle won't. There is possibly a way you can do it which is to put 2 sockets at 
each point and use a jumper cable between them, but any overhang will cause massive reflections and 
horrible mismatches.

X---X=X--X=X---X=X---X

Where X is a socket, - is the structured cable and = is a jumper will work, but doing it the way you 
are doing it will not ever be reliable even if you can get it to work.

Maybe.. and it's a BIG maybe if you built a little terminator with fixed resistors of the right 
value to plug into the end socket you might get it to be reliable enough, but what you are doing 
violates every known standard.

I'd be putting a pair of sockets at each point and using small jumper cables when you need to 
connect them together. This also violates the specs with regard to number of terminations, but it'll 
most likely work.

Brad
-- 
"Human beings, who are almost unique in having the ability
to learn from the experience of others, are also remarkable
for their apparent disinclination to do so." -- Douglas Adams