Talk:Twisted pair

From Wikipedia, the free encyclopedia

Regarding unshielded twisted pair

If it ain't shielded, it'll short-circuit. So someone should correct this unprecise phrase, please. thanks, --Abdull 12:08, 17 Mar 2005 (UTC)

you mean insulated. electromagnetic shielding is different. - Omegatron

"f cancelling out electromagnetic interference kno"

technically magnetic interference, right? only shielding cancels electric interference. - Omegatron 16:55, Apr 15, 2005 (UTC)
Hmm.. effect of the magnetic field from one wire on the electrons from the other thus electromagnetic?? Mozzerati 19:42, 27 September 2005 (UTC)
Well I don't think the method is efficient enough to cancel out direct magnetic interference, so the real thing it cancels out is crosstalk, which actually is EMI


Well, actually, electric and magnetic fields are inseparable. You can't just have a magnetic field or just an electric field. They're orthogonally oriented components of each other. So both are probably canceled by the twisting.

When the wavelength is much longer than the wire length they are pretty much separate, and when it is much longer than the wire spacing they can still be treated separately.

Consider the electromagnetic field due to a nearby radio station. The induced voltage will be pretty much the same on both wires, called common-mode. That won't get through the transformer at the end.

Consider a 50Hz or 60Hz magnetic field from a nearby power cable. An untwisted cable loops like a very long and narrow wire loop, and a changing magnetic field will induce a current. Twisting with a pitch somewhat less than the wavelength causes the induced currents to change sign every half twist.

Consider two twisted pairs in the same cable. If the twist pitch is about equal, both inductive (magnetic) and capacitive (electrostatic) coupling can add over the length of the cable. (Model capacitive coupling as capacitors between the wires, larger capacitors for closer wires.) With different twist pitch, the induced current and voltage cancels over lengths somewhat longer than the shorter twist pitch.

Gah4 23:25, 22 June 2007 (UTC)


Contents

[edit] Images for the article

UTP cable
UTP cable
Image:S-UTP-cable.png
FTP cable
STP cable
STP cable
S/STP cable
S/STP cable

Hi I translated the images from the German article into english, please check wether I got that right before introducing into the article. --Deelkar (talk) 01:13, 15 Jun 2005 (UTC)

Looks good! "Mantle" should be "sheath", though.
http://www.hifi.org.uk/tech/images/buc.gif - Omegatron 01:28, Jun 15, 2005 (UTC)
Changing it now. Thanks! --Deelkar (talk) 02:01, 15 Jun 2005 (UTC)

There are contradictions between the German and the English article concerning the nomenclature for STP, S/FTP and S/UTP. Especially because some manufacturers seem to call their cable STP while meaning S/UTP.--Deelkar (talk) 02:26, 15 Jun 2005 (UTC)

Are there any formal references that clearly says what is STP/ScTP/FTP, etc. Unlike in this article, this is how Cisco defines it:
STP 
Screen around each pair and the whole shebang! (S/STP in this article)
FTP ; ScTP 
Synonyms. Outer foil or metal net, but no screen for individual pairs.
UTP 
No shoielding at all.--itpastorn (talk) 17:33, 2 March 2008 (UTC)

[edit] need of help

This page may need the expertise of a professionalAkupta321 06:31, 18 March 2006 (UTC)

[edit] Merge stubs into this article

I have already gone ahead and merged the shielded twisted pair stub into this article. I would also like to do so with the unshielded twisted pair article. Before I do so, I'd like some other people's opinions. N. Harmon 01:35, 28 May 2006 (UTC)

I think the foiled twisted pair article should be merged into here also. N. Harmon 01:43, 28 May 2006 (UTC)

I also think we should consider merging several cable category articles into this one. N. Harmon 02:22, 28 May 2006 (UTC)

Oppose the merger of cable category articles, but support the merger of the twisted pair variants. The cable category articles are (from what I can tell) pretty good, this article shouldn't concern itself with them. By the way, from the history I noticed that the user Mozzerati originally seperated the sections, see what he has to say. Also, take a look at all the articles pointing here and see if it is possible to improve that. In my opinion this article is in need of some serious attention, I'll help around where I can. --Bruce 09:54, 28 May 2006 (UTC)
Two of the cable category articles are good, specifically Category 5 cable and Category 6 cable. However they do seem to repeat a lot of information that is really about twisted pair cables and not necessarily THAT category of cable. I appreciate your thoughts on this. N. Harmon 15:37, 28 May 2006 (UTC)
Oppose keep specialized info in seperate articles. Generic info might be merged into twisted pair. Bemoeial 12:48, 6 June 2006 (UTC)

"'oppose'" - There are hundreds of type of twisted pair cables, but "Category X" is not a generic term, it is a specific term for this particular cable. It deserves a place of it's own. Trane Engineer

"'oppose'" The information for Cat 6 cable was good, and should be in its own catagory. <pun> I would not look it up under twisted pair, I would search under Category X. Twisted pair is too general, like putting it all under "wire". 69.31.216.194 17:48, 6 June 2006 (UTC) Tom Moffatt

"'Oppose" It does take some knowledge of most types of media cable to differenciate twisted pair. It can't just be classified to just Cat5 cables! ~ T. Hansen, California

  • Oppose Too much info to merge, Wikipedia is not paper. Themindset 16:15, 9 June 2006 (UTC)
  • gently oppose there's a real need for a discussion of the physics, grounding characteristics etc. of the different types. I started to think about writing it, but got sidetracked. I think they are sufficiently important and different to be worthy of separate articles. On the other hand, there's nothing wrong with merging for some time and then later splitting again. Mozzerati 20:31, 9 June 2006 (UTC)

[edit] Cancelation between pairs

To me it is not obviously why two pairs in the same cable does not crosstalk a lot. It looks like they are two coils on the same transformer. I expect that they are twisted at the same rate. Appearently this is not the case and I think this article should explain why. 62.79.30.244 08:23, 12 February 2007 (UTC)

  • "Why does each pair in Cat 5/5E/6 cable have different twist rates?" [1]
  • "EIA-568 ... Pair Assembly: The pair twists of any pair shall not be exactly the same as any other pair. The pair twist lengths shall be selected by the manufacturer to assure compliance with the crosstalk requirements of this standard." [2]
  • "... within the Cat 5 cable. Each individual phone line ... each phone line (made of two twisted strands of copper) has a "twist rate" different from the other three lines. One of the four pairs will have a high number of twists per inch, one will have a low number of twists per inch, and the other two will have twist rates somewhere in between. That difference in twist rates prevents signals from bleeding across phone lines." [3]
  • "UTP (Unshielded Twisted Pair) cable rated as Category-5 typically contain (8 conductors) 4 pairs of 24 awg solid copper wire. Each of the pairs is two conductors twisted together, and each pair in a cable has a different number of twists per inch when compared to the other pairs." [4]
  • "CAT5/5e/6 cables consist of four pairs of cables... the slight difference in twist rate for each pair... " [5]
  • twist rate pedantry [6]

--76.209.28.72 18:34, 5 June 2007 (UTC)

To simplify what the anon above has stated: Take a spare cable that you have, and remove the shielding; notice how each of the four pairs is twisted differently, with each pair having a different number of twists? Since the rate of twists for each pair is different, crosstalk (the aforementioned "bleeding") is less likely to occur. PeanutCheeseBar 19:24, 14 June 2007 (UTC)

[edit] differential vs balanced

I'm just trying to clear up misconceptions around the 'pedia about EMI and cables. As I understand it:

  • "Balanced" refers to keeping the source and load impedances of a transmission line equal, so that any noise that couples into the wires will couple equally, and can be rejected at the other end.
  • Twisting cancels out fields that couple into adjacent twists.
  • "Differential" refers to sending signals of opposite polarity down the wires. This is so that the wire's EM fields mostly cancel out.
  • Shielding does a variety of things

— Omegatron 16:05, 31 May 2007 (UTC)

Balanced and differential are essentially the same. Keeping the source and load impedance the same is matched, usually impedance matched.

Balanced means the currents (and for uniform impedance, the voltage) are equal and opposite in the two wires. Both electric and magnetic fields fall off fairly fast with distance

Yes, it is really the electromagentic field, but one can separately consider capacitive coupling (due to the electric field) and inductive coupling (due to the magnetic field). Two parallel UTP cables with the same twist pitch will have both capacitive and inductive coupling as the fields add up over the length of the cable. With different twist pitch, in the long run the coupling, both inductive and capacitive coupling average out to zero.

Note that impedance matching is not required for UTP operation. Assuming transformers are used on both ends, the reflected signal will still be balanced.

Gah4 Gah4

  • Yes, matching source and load impedances is another important feature, to prevent reflections and "ringing".
  • "Balanced" applies to the impedances of each line, though, not the currents or the signal. It means that both lines are driven by the same source impedance and terminated by the same load impedance (and have the same characteristic impedance?), so that any noise that couples into them will be equal in both and can be rejected. — Omegatron 23:32, 22 June 2007 (UTC)

A reference to balanced is probably applicable. It looks pretty good, but is mostly meant for low frequency (audio) applications. UTP works just find if the impedances are wrong, as long as they aren't too far off. If transformer driven, the currents will be equal and opposite, which is most important as far as EMI (radiation) is concerned. Equal impedance to ground is important, but is reasonably likely in a twisted pair cable.

Note that telephone wiring uses 600 ohm source/load but the UTP cable has much lower impedance. Yes, the impedance for each wire to ground must be (approximately) equal, but that is fairly unrelated to the cable impedance. If the impedances to ground are different then equal voltage/current in won't stay equal.

Gah4 08:08, 10 September 2007 (UTC)

I added the Historical Note, forgot to enter Edit Summary.LoopTel 19:23, 5 July 2007 (UTC)

[edit] Disagreement between advantages/disadvantages

First we say that "It is a thin, flexible cable that is easy to string between walls." Then we say that "...twisted pair cables usually have stringent requirements for maximum pulling tension as well as minimum bend radius. This relative fragility of twisted pair cables makes the installation practices an important part of ensuring the cable’s performance." Which is it? — BRIAN0918 • 2007-09-12 15:11Z


[edit] Mains twisted pair

The article appears to have not noticed that there is also twisted pair flex that has nothing to do with data transmission, and is twisted solely because its the lowest cost way to keep 2 insulated conductors together. Such flex was in widespread use in the first world until the 1960s as mains flex, and is still sometimes found in poorer countries. Tabby (talk) 12:13, 1 March 2008 (UTC)

[edit] Why UTP

"Unshielded twisted pair (UTP) cabling, because of its 100-year history of use by telephone systems, both indoors and out, is also the most common cable used in computer networking."

no, because its the cheapest thing that does the job. Thats more or less always the reason for such decisions. Tabby (talk) 12:13, 1 March 2008 (UTC)

by (http://tldp.org/HOWTO/Serial-HOWTO-5.html#ss5.2) twisted pair technology obtains higher speed Vy0123 (talk) 04:18, 19 March 2008 (UTC)

Compared to the serial port - does not mean much! Tabby has already said all there is to say about this subject.--itpastorn (talk) 09:26, 19 March 2008 (UTC)

Tabby refers to price. The tdlp refers to the principle of twisted pair technology in writing:-

All of the EIA-232 serial cable wires use a common ground return wire so that twisted-pair technology (needed for high speeds) can't be used without additional hardware.

It is somewhat tragic that the RS-232 standard from 1969 did not use twisted pair technology which could operate about a hundred times faster. Twisted pairs have been used in telephone cables since the late 1800's. In 1888 (over 115 years ago) the "Cable Conference" reported its support of twisted-pair (for telephone systems) and pointed out its advantages. But over 80 years after this approval by the "Cable Conference", RS-232 failed to utilize it.

Is it possible that the value of the principle of twisted pair technology can lead to enduring use and ubiquity and, in the process, economies of scale in production draw down the price of utp cabling? Vy0123 (talk) 11:27, 19 March 2008 (UTC)

[edit] This needs dumbing-down to make it more accessible

I read this article because I have read that twisted pairs are a modern alternative to coaxial cable for data transmission on local networks. Specifically, I wanted to understand what goes on in the "pairs". I couldn't get my answer from the article - it assumes I already know the answer. "two conductors are wound together for the purposes of canceling out electromagnetic interference (EMI) from external sources... In balanced pair operation, the two wires typically carry equal and opposite signals (differential mode) which are combined by addition at the destination". Are the two wires in each pair the quivalent of the two wires that provide the circuit powering a light bulb or other device ? Do data lines require 2 wires the same way power lines do ? What does it mean, and how does it happen, that the two wires in a pair carry equal and opposite signals ? Is a single twisted pair equivalent to a coaxial cable ? Or does each twisted pair provide a single electric wire, and two twisted pairs are required to provide a complete circuit ? Please consider making the article comprehensible to the intelligent reader who knows nothing about the subject. Rcbutcher (talk) 18:10, 22 May 2008 (UTC)

Excellent observation! To answer your main question: A twisted pair is two halves of one circuit. Hopefully, my quick addition to the lead will help start the simplifying process. (Rather than "dumbing down" which sounds like removal of difficult material, I prefer to think that the article will be expanded downward to include readers who are beginners to the topic.) Binksternet (talk) 01:09, 23 May 2008 (UTC)
Sounds good. Indeed, what I meant was not to remove difficult material but to make it more accessible to the mentally active reader who knows very little about the subject but wants to get a basic accurate understanding of it, in this case how the wires going from his PC's LAN card to the hub or cable modem work. Rcbutcher (talk) 14:33, 23 May 2008 (UTC)