My answer is at the end of the quoted text (normally I don't
quote everything back, to save space) (since my reply
usually is just to one point, andthe rest then not relevant)
----- Original Message -----
From: "Paul M Foster" <paulf@quillandmouse.com>
To: <slug@nks.net>
Sent: Wednesday, July 09, 2003 23:43
Subject: Re: [SLUG] cable 568b wiring --was: brighthouse
> On Wed, Jul 09, 2003 at 01:17:36AM -0400, Robert Foxworth wrote:
>
> >
> >
> > Here is the T-568B wiring spec, recommended for network
> > wiring (not telco wiring which is where -A or USOC is used)
> >
> > The only difference is that the wire-pin assignment for
> > pairs 2 and 3 are reversed. If you are consistent, it makes
> > no electrical difference. The info below was done with
> > monospaced courier and tabs, btw.
> >
> > Straight through: same format at BOTH ends:
> >
> > Pin# Pair# Function Wire Color
> > 1 2 Tx+ white/orange
> > 2 2 Tx- orange
> > 3 3 Rx+ white/green
> > 4 1 --- blue
> > 5 1 --- white/blue
> > 6 3 Rx- green
> > 7 4 --- white/brown
> > 8 4 --- brown
> >
> > As you see, only pairs 2 and 3 conduct signal.
> > Pairs 1 and 4 are electrically idle, but the spec
> > says they are all connected. When making cables you
> > need to connect them anyway to maintain proper
> > wire position.
> >
> > To make a crossover cable, pairs 2 and 3 are
> > reversed at ONE end. Use the above format at
> > the first end, and use T-568A at the second,
> > as below:
> >
> > pin# pair# function wire color
> > 1 3 Tx white/green
> > 2 3 Tx green
> > 3 2 Rx white/orange
> > 4 1 --- blue
> > 5 1 --- white/blue
> > 6 2 Rx orange
> > 7 4 --- white/brown
> > 8 4 --- brown
> >
> > When looking into the mouth of the RJ45 jack,
> > with the contact wires up and the lock tab
> > down, pin 1 is on the (your) left.
> >
>
> This is a _great_ description. I have a question, though. I'm not sure I
> understand your pair designations. Based on your description, I assume
> that the pairs are numbered this way:
>
> Pins: 8 7 6 5 4 3 2 1
> . . . . . . . .
> Pairs: 4 4 2 1 1 2 3 3
>
> Is that right? I know that RJ11 jacks number the first pair as the
> inside wires and the outside wires as pair two. But is the above the way
> pairs are numbered when 8 conductors are involved, regardless of what
> type of wiring scheme is being used?
>
> Paul
I think I understand the disconnect in thinking here, is to assume that
each of the 4 pairs should have been installed consecutively on
the 8 pins in the jack. In fact, only 3 of the 4 pairs are attached to
consecutively numbered pins.
I suppose this is a good illustration of the problem that authors and
doco writers have in expressing concepts verbally, and I admit that I
myself could have done better. One great treatment of this problem
(expressing pinouts) is found in the MilesTek catalog (a supplier of
hardware, cables etc. 1-800-524-7444, ask for a catalog). Their
2003-6 catalog has a pictorial of the pin connectors. Why every other
book author can't do this, I jus' dunno. But when you see it, you'll
find that the spec uses three of the 4 pairs on the two outermost
available pins, and on the two center pins. That leaves pins 3 and 6
unassigned. The last pair then _straddles_ the middle pair. The
difference then, is that in T568A pairs 3,1 and 4 are each wired on
adjacent pins, and pair 2 straddles pair 1. In the case of T568-B
then, pins 2,1 and 4 are on adjacent pins and pair 3 straddles pair
1, the straddling pair being on pins 3 and 6 in each case.
I assume that this done so that telco battery voltage, tokenring etc.
won't cause problems when accidentally connected to ethernet
jacks since the physical media is similar in such cases. Since
the straddling pair, and the first end-pair are the active pairs in
10baseT, and the other schemes use center-adjacent pairs, this
helps make it luser-proof, I suppose.
HTH. Bob (the unrealized tech-writer angst welling up in glorious
rebellion against book authors everywhere for not dealing with this
more clearly...?)
(say, Hillary should have included this, then her book would not then
have been a work of fiction)
This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 19:03:37 EDT