Searching \ for '[EE]: Cheapskate network cabling' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=cheapskate+network
Search entire site for: 'Cheapskate network cabling'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: Cheapskate network cabling'
2002\05\01@213547 by Russell McMahon

face
flavicon
face
Questions:    How reliably are Cat5e network cables when used
                   to carry two 100 Mbps signals simultaneously using
                  the two "spare" pairs?

                What are the affects of running other low speed or
                analogue circuits on the unused pairs?

Discussion:        100 Mbps Cat5e network cables have 4 pairs in the sheath
but use only two. There are two standard wiring versions  (A & B) (one uses
green & orange pairs, other = ?). Only difference is which pairs are used. I
expect that the cable design and the standard take account of the relative
location of the pairs within the "quad" formed by the 4 pairs and attempt to
maximise balance and minimise crosstalk.

The version of the standard which I am using is wired
green/grnwht/org/bluwht/blue/orgwht/brown/brnwhite
I have noticed that flexible cables sold commercially almost invariably
DON'T adhere to either standard. They are usually wired 'straight" eg
grn/grnwhite/org/orgwhite/blue/bluwhite/brown./brnwhite. Assuming the NIC is
connected in a manner which anticipates the cable being wired to the
standard, this has the result of "splitting" the second pair used onto a
separate leg of two physical pairs. This is an abominable thing to do and
goes against every form of good practice known - even at audio frequencies,
let alone at 100 Mbps. Despite this a 30 metre flexible cable seems to work
OK at 100 Mbps !.
Clearly there is some margin for error. I would expect severe impedance
bumps at both ends of the cable and a degradation of the BER. The switch I
am using gives a visible indication when it drops back to 10 Mbps and it is
not doing so with these cables. A proper tester, which I have not got, or a
TDR, which I also have not got, would give some indication of what was
really happening.

Some cables I have seen have only two pairs used and these are wired to pins
1,2,3,6 which is what is required by the standard. Clearly the NICs ARE
expecting this arrangement. This cable also works at 100 Mbps without
"apparent" problems.

This obvious tolerance to pathetic practice encourages me to wonder if I can
RELIABLY run two 100 Mbps circuits on a single CAT5E cable. eg green &
orange for one circuit (as per standard) and blue and brown for the other. I
can of course easily try this (and will) but even if it works (as I'm sure
it will) I will not have a good idea of the effect on BER without extensive
testing.

I have also heard of the other pairs being used for eg telephone circuits
and I'm sure that would also work "well enough"
I have two CAT5E cables going to required locations in my premises. These
use flexible cable. Adopting various of the above abominable practices would
allow greater capacity traded off against unknown degradation.

SO -

Has anyone got any practical experience or knowledge about the effects of
running two x 100 Mbps circuits in one cable and/or other lower speed
circuits?

--
http://www.piclist.com hint: To leave the PICList
spam_OUTpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu


2002\05\01@215805 by Mitch Miller

flavicon
face
On Thu, 2 May 2002, Russell McMahon wrote:

> Questions:    How reliably are Cat5e network cables when used
>                     to carry two 100 Mbps signals simultaneously using
>                    the two "spare" pairs?
>
>                  What are the affects of running other low speed or
>                  analogue circuits on the unused pairs?

I can't say I've specifically tried using the other two pairs for data
(nor anything for that matter), but I do know that the pairs are twisted
at different rates.  It seems like the brown pair is always twisted much
looser than the others.  Not sure if that would affect performance or not.

-- Mitch

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam@spam@mitvma.mit.edu


2002\05\01@222725 by Russell McMahon

face
flavicon
face
> > Questions:    How reliably are Cat5e network cables when used
> >                     to carry two 100 Mbps signals simultaneously using
> >                    the two "spare" pairs?
> >
> >                  What are the affects of running other low speed or
> >                  analogue circuits on the unused pairs?
>
> I can't say I've specifically tried using the other two pairs for data
> (nor anything for that matter), but I do know that the pairs are twisted
> at different rates.  It seems like the brown pair is always twisted much
> looser than the others.  Not sure if that would affect performance or not.


Such a revelation demanded immediate testing. And you are, of course,
correct. It's obvious after the event and I'm sure it's no surprise to those
who are intimately involved in such things but ALL the twist rates are
different !!!
This is no doubt designed to minimise crosstalk and I imagine that the
repetition ratios are aimed at being as close to prime relative to each
other so that you never get a repeating pattern.

This means that the "pair splitting:  achieved by careless commercial cable
makers is even worse than I expected as the two legs are not only in
different pairs but are of different wire lengths die to the different pitch
of the twist (unless the twist rate varies along the cable's length, which
is possible but not likely).

Pitches were APPROXIMATELY (despite significant digits shown :-) )

   Orange        1.55 mm
   Green          1.75 mm
   Brown         2.2mm
   Blue            2.65 mm



           Russell McMahon

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestspamKILLspammitvma.mit.edu


2002\05\01@224046 by Lee Jones

flavicon
face
> Questions:    How reliably are Cat5e network cables when used to
>        carry two 100 Mbps signals simultaneously using the two
>        "spare" pairs?

>        What are the affects of running other low speed or analogue
>        circuits on the unused pairs?

> Discussion: 100 Mbps Cat5e network cables have 4 pairs in the
> sheath but use only two. There are two standard wiring versions
> (A & B) (one uses green & orange pairs, other = ?).

Only difference between T568A and T568B is location in the RJ45
connector (plug or jack) of white-green and white-orange pairs.o

They are as follows:

  pin       T568A            T568B

   1     white-green      white-orange   <-- Ethernet transmit
   2     green-white      orange-white   <-- Ethernet transmit
   3     white-orange     white-green    <-- Ethernet receive
   4     blue-white       blue-white

   5     white-blue       white-blue
   6     orange-white     green-white    <-- Ethernet receive
   7     white-brown      white-brown
   8     brown-white      brown-white


> Only difference is which pairs are used.

Nope, same pairs are used.  It just differs in which pair is used
for transmit and which for receive.  (Obviously, this is relative
to a consistant reference such as the device end; the hub end will
be flipped assuming a straight through cable.)

FDDI on copper used white-green (pins 1 & 2) and white-brown (pins
7 & 8).  But FDDI lost the market to 100 Mbps Ethernet.

> I expect that the cable design and the standard take account of
> the relative location of the pairs within the "quad" formed by
> the 4 pairs and attempt to maximise balance and minimise crosstalk.

In real life, the position of individual pairs within the outer
sheath varies over long runs.  There's also 25 pair cable rated
Category 5, so the pair relationships can't hold there.

> The version of the standard which I am using is wired
> green/grnwht/org/bluwht/blue/orgwht/brown/brnwhite

That's T568A.  As long as both ends of each pair are swapped, it
works fine (and you don't have polarity issues (which almost all
network adapters correct automatically)).

> I have noticed that flexible cables sold commercially almost
> invariably DON'T adhere to either standard. They are usually
> wired 'straight" eg grn/grnwht/org/orgwht/blu/bluwht/brn/brnwht.

Cables wired this way cannot be sold as Category 5.  If I received
any of these from a vendor, they would go back with a nasty gram.

> This is an abominable thing to do and goes against every form
> of good practice known - even at audio frequencies, let alone
> at 100 Mbps. Despite this a 30 metre flexible cable seems to
> work OK at 100 Mbps !.

I've encountered this before.  Our equipment was located at a
tradeshow and was getting high error rates.  I had the cable
replaced with one wired correctly and the problems vanished.

> Some cables I have seen have only two pairs used and these are
> wired to pins 1,2,3,6 which is what is required by the standard.
> Clearly the NICs ARE expecting this arrangement. This cable
> also works at 100 Mbps without "apparent" problems.

As long the wires on pins 1 & 2 are paired (twisted together) and
the wires on pins 3 & 6 are paired (twisted at a different rate),
the cable will work fine.

> This obvious tolerance to pathetic practice

Ethernet has _lots_ of margin built in.  I think this is why it
has done so well in the marketplace.

> encourages me to wonder if I can RELIABLY run two 100 Mbps
> circuits on a single CAT5E cable. eg green & orange for one
> circuit (as per standard) and blue and brown for the other.

> Has anyone got any practical experience or knowledge about the
> effects of running two x 100 Mbps circuits in one cable and/or
> other lower speed circuits?

I've always wondered this myself.  I do have one site which has
it's satellite distribution frames (SDFs) connected back to the
main distribution frame (MDF) via 25 pair Category 5 cables.  It
works fine with many, many 10 and 100 Mbps Ethernet drops at the
far ends.  Now the pairs have more physical room inside a 25 pair
cable sheath, so specific pairs are as close to each other for
long distances -- but randomness may cause it to be so.

I've always had in-mind that I could put two Ethernet drops on
a single 4 pair Cat 5 cable -- if I _REALLY_ had to -- but I've
always put in enough wire that I haven't had to stoop to it.

I'm interested in what data you collect on dual Ethernet drops
on a single Cat 5, 5e, or 6 cable.

> I have also heard of the other pairs being used for eg telephone
> circuits and I'm sure that would also work "well enough"

Blue & blue-white are in the center of the RJ45 jack so they line
up with an RJ11 plug (to an analog phone) inserted into the jack.
Sharing phone with Ethernet does work with no noticable problems.

                                               Lee Jones

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam.....mitvma.mit.edu


2002\05\01@230714 by Bob Ammerman

picon face
My inexpensive commercially made cables are wired correctly. The pairs are
interleaved properly on the connector.

The two standards for pinouts are called XXX486A and XXX486B where XXX is
some magic three letter acronym, (TIA perhaps IIRC?). I believe that alnost
all data (as opposed to voice) wiring is done consistenty with one of the
standards (but don't ask me which one).

I have used CAT5 (not even 5E) to send two simultaneous 10Mbit Ethernet
signals at once without trouble.

Also, remember that CAT5 supports 100MBit full duplex. This implies a
certain amount of immunity to what is happening on at least one of the other
pairs. I would guess you would have pretty good results.

Bob Ammerman
RAm Systems

--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspam_OUTspamTakeThisOuTmitvma.mit.edu


2002\05\01@231854 by A.J. Tufgar

flavicon
face
I used to run two ethernet 100Mbps all the time off the same wire.  80
meters through a factory (large inductive loads starting and stoping)
was the max I ever tried though (with multiple drops).  This ran fine
with 0% packet loss.  I believe that the other two pairs were suppose to
be used as telephone pairs to run on the same cable, this is how I have
often seen it done in larger offices.

If you want dirt cheap, you can get away with normal phone line 2 pair
straight through cable (stuff telephones use), although I wouldn't go
more then 50 meters and not through a noisy enviroment.

Never really had an episode of bad cabling, usually I'd recrimp the end
if I expected trouble and it would work fine.

All and all worked fine for me,
Aaron

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestspamspam_OUTmitvma.mit.edu


2002\05\02@005730 by M. Adam Davis

flavicon
face
It works fine.  The reason it's not common is because the jacks are
wired for future ethernet standards that may use the other two pairs.
Some early 100Mb gear did use all four pairs.  I've heard that 1Gb gear
does use all four, but I don't remember where and could easily be wrong.

You can buy ethernet splitters that does exactly this.  It has two jacks
and one plug.  You buy two and put one one one end and one on the other
- it eliminates having to rewire the jacks in the office.

There are many devices out there that use the extra pairs for power as
well.  Used for ethernet cameras, remote powerless hubs, some 802.11
access points, etc.  Not terribly common, I suspect becuase there is no
standard and they are often incompatible from manufacturer to manufacturer.

-Adam

Russell McMahon wrote:

{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\05\02@040152 by Bond, Peter

flavicon
face
> From: Russell McMahon

> Questions:    How reliably are Cat5e network cables when used
>                     to carry two 100 Mbps signals simultaneously using
>                    the two "spare" pairs?

www.videkonline.co.uk/SubSections.asp?sectionID=22#
I imagine this means they work acceptably well - but how high a BER is too
high?

>                  What are the affects of running other low speed or
>                  analogue circuits on the unused pairs?

No idea.  My father was interested in running CH thermostat signals on the
other pairs, but hasn't done any checks yet to see if it compromises data
integrity.

Peter
This email, its content and any attachments is PRIVATE AND CONFIDENTIAL to
TANDBERG Television. If received in error please notify the sender and
destroy the original message and attachments.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\05\02@045641 by michael brown

flavicon
face
> If you want dirt cheap, you can get away with normal phone line 2 pair
> straight through cable (stuff telephones use), although I wouldn't go
> more then 50 meters and not through a noisy enviroment.

ATTENTION:  Never do this.  Silver satin will cause innumerable problems
even if used as 6' patch cords.  The funny thing about this is that IPX will
work slowly, TCP/IP will cause windoze blue screens and huge delays due to
the excessive errors.  I spent/wasted hours at a customers site due to this.
They had used silver satin as patch cords and were using a hub (not a
switch).  I was replacing a croaked Novell server with a Linux box.  A truly
incredible (just like Green Acres) set of symptoms.  This was in a sign
factory with relatively little EMI.  No cabling went anywhere near the shop
area.  It seems that MS TCP/IP can only handle a certain error rate before
keeling over.  I found all the silver patch cords and switched them all out
with "real" patch cords and the problems instantly went away (no other
changes).

I have used the extra pairs (brown and blue) to carry telephone signals with
no apparant problems, although the CAT-5 spec. clearly indicates that all 4
pairs should be terminated in the RJ45 connector.  100VG networks do use all
4 pairs.

michael brown

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\05\02@071710 by Olin Lathrop

face picon face
> There are many devices out there that use the extra pairs for power as
> well.  Used for ethernet cameras, remote powerless hubs, some 802.11
> access points, etc.  Not terribly common, I suspect becuase there is no
> standard and they are often incompatible from manufacturer to
manufacturer.

Such a standard is being worked on now, although I think is uses the
existing two pairs.  I think a draft is already available.  You will soon
see hubs that supply power over the ethernet and low end ethernet devices
that require such power.  Kinda like USB.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\05\03@053900 by Amaury Jacquot

flavicon
face
Quoting Olin Lathrop <KILLspamolin_piclistKILLspamspamEMBEDINC.COM>:

> > There are many devices out there that use the extra pairs for power as
> > well.  Used for ethernet cameras, remote powerless hubs, some 802.11
> > access points, etc.  Not terribly common, I suspect becuase there is no
> > standard and they are often incompatible from manufacturer to
> manufacturer.
>
> Such a standard is being worked on now, although I think is uses the
> existing two pairs.  I think a draft is already available.  You will soon
> see hubs that supply power over the ethernet and low end ethernet devices
> that require such power.  Kinda like USB.

This is for example used to power Cisco IP phones from the switch.
(I know, as I have a shitload of them here in the lab)

Sincerely

Amaury

{Quote hidden}

--

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...