Talk:Category 5 cable/Archive 1
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 |
Insulation type
I could be wrong, I have not seen the spec, but I believe that there is a critical piece of magic in the choice of the insulation of the wires. I think that FEP teflon is used because it has a low dispersion, and that this is critical to extending the useful distance that these rather wideband signals can propagate. I am sure they would use something cheaper if they could. Clearly, the teflon is not used for fire resistance, since the jacket is often vinyl. Dispersion is sometimes called delay distortion in this realm. It occurs when the index of refraction (or dielectric constant) changes over the frequency range of interest, causing pulse-like waveforms to spread out in time. If this is true, it might be fun to mention it. AJim 23:13, 16 Mar 2004 (UTC)
- AJim, that's an interesting comment, but the rating of the cable deals with it's sheathing which - of course - plays no role as a dielectric. Teflon is chosen for it's high smoke point and lack of noxious gases. Regular cable uses PVC which has a low smoke point and emits noxious fumes. Teflon sheathed cable has the added benefit of not needing "cable lube" when running through conduit although no one uses it purely for that attribute due to cost. Cataclysmcow 1:51, 08 May 2005 (MST)
Distance limit
The statement that Cat5e is limited to "350 m, compared with 100 m for Cat5" in relation to 100Base-T is incorrect. The effective distance of any wiring plan is dependant on the signaling scheme used. In ethernet the maximum length is dependant on the propagation delay of the medium. This propagation delay remains mostly constant for all UTP. The effective length of ethernet over any UTP is 100m. Cataclysmcow 8:58, 08 May 2005 (UTC)
- isn't propogation delay only an issue for the half duplex variants of ethernet? Plugwash 21:01, 17 July 2005 (UTC)
- A quick search turns up numerous sources claiming the maximum length to be 100m (328 feet) for category 5 and 5E twisted pair. I think perhaps someone got confused with Cat-5E operating/rated at 350MHz. --Southen 06:51, 24 September 2005 (UTC)
- Moreover, cat5e, cat6, cat7 and very likely whatever there is to come can not guarantee distances over 100m for 100Mbit links. All higher categories are backward-compatible, but chances of getting 100Mb working over 100m+ links can actually be diminished by upgrading to cat6/7. The 350m statement should be erased from the article and the opposite should be added. I'll do it later if no one steps forward before.
- A quick search turns up numerous sources claiming the maximum length to be 100m (328 feet) for category 5 and 5E twisted pair. I think perhaps someone got confused with Cat-5E operating/rated at 350MHz. --Southen 06:51, 24 September 2005 (UTC)
CAT5/ CAT5e
So you cannot take Cat5 or CAT5e over 100m? How is the VoP of a certain ethernet system figured, For example; if I needed to run CAT5 of CAT5e up 16 floors, or 160ft and across 160ft that makes 320ft. Is that possible? If I can find the VoP of the system then CAT5e should be able to run the distance as I understand it??? Otherwise I am stuck setting up a new T1 line and then having to set up everything else that comes with stand free network. Any ideas and suggestions would be much appreciated. {{Subst:unsigned|
- use one of the fiber variants of ethernet? while they are far more expensive than CAT5 for a link between cabinets it shouldn't be prohibitively expensive. Plugwash 21:00, 17 July 2005 (UTC)
- Wiki is not the right place for you to find solution. Anyway, for your case, you can either use a repeater to extend the signal, or use another switch to extend the signals. —Preceding unsigned comment added by 219.93.185.42 (talk • contribs)
568A v B
According to the TIA 568A and B page, both are electrically equivalent. From this I judge that the section which describes one as straight-through and one as crossover is incorrect. Is this correct? --69.207.105.212 03:23, 11 October 2005 (UTC)
- Yes. There is absolutely no functional difference between the two. A straight-through cable has the same type (A or B) on both ends and a crossover has A on one end and B on the other.--Roofus 06:53, 11 October 2005 (UTC)
- I rewrote the section. It should be more accurate, even if it's not particularly eloquent. :P--Roofus 07:02, 11 October 2005 (UTC)
I think there is another A versus B ambiguity here. From another WP page (TIA/EIA-568-B), I quote:
"The three standards are formally titled - ANSI/TIA/EIA-568-B.1-2001, -B.2-2001, and -B.3-2001." [Ed. you see that these are all "B" titles.] "The TIA/EIA-568-B standards were first published in 2001. They supersede the TIA/EIA-568-A standards set, which are now obsolete." Pin assignments are named T568A and T568B, and are frequently referred to (erroneously) as TIA/EIA-568A and TIA/EIA-568B.
I think that the diagrams do not have the right titles, am I right? Or is the other page wrong? (I'm new here, so sorry if this doesn't make sense) —Preceding unsigned comment added by 81.158.150.141 (talk) 22:45, 30 December 2008 (UTC)
Cabling guide
In the external links section there is a link to http://www.linux.ie/articles/tutorials/networking.php. That page was written in 1999 (over 6 years ago now), and presents 10Base-T as a new and better choice (hellooo? 100Base-T and GbE anybody? 3rd millennium?)
Is there any reason why this link should be listed here instead of a "history of networking" page?
Aditsu 07:04, 15 February 2006 (UTC)
- It's still got some good information on it. The cabling procedure hasn't changed. --Roofus 15:06, 15 February 2006 (UTC)
- What has changed is the care you must take to get the cabling right if you wan't the faster speeds (especially gigabit) to work properly. Plugwash 14:02, 12 June 2006 (UTC)
Not funtioning
cat 5 cable, i was used about 50 metes for my VOIP video phone but its not functioning ....? —Preceding unsigned comment added by 210.19.204.210 (talk • contribs)
- Most likely causes are bad connection damaged cable and split pairs. after checking it is wired correctly try re-terminating it using the other two pairs as the signal carrying ones. Plugwash 19:56, 20 February 2007 (UTC)
What is the Ripcord inside most cat5e cable for?
No one has discussed this topic, or even mentioned that cat5e cable usually has a ripcord inside in addition to the copper wires. What is this ripcord for? To tear the cable open lengthwise? —Preceding unsigned comment added by 131.212.109.27 (talk • contribs)
- Yes, that's exactly what it is for. dpotter 23:59, 15 September 2006 (UTC)
Define cat5e Ripcord: When cutting the outside insulation of cat5e in order to expose the inside wires, it is common to nick/cut the insulation on the wires prior to crimping an attaching connector. The "Ripcord" allows the inside cord (not wire) to be pulled through the insulation to expose the rest of the wire w/out the risk of random cuts. Todays connectors rely on the insulation to be intact rather than "stripped" off in order to hold a connection in place. The "Ripcord" idea is to manually pull down the string to clear insulation with zero risk of exposing open wire to a short from cut insulation.
I must now say that a small cut before attempting to bend the plastic insulation to it's breaking/cut point is timesaving. That, or strip a good handful of wire before using the "Ripcord." Rip cord is good. 68.218.61.120 09:47, 8 February 2007 (UTC)
I thought the fiber strand was to keep the cable from being stretched and changing the characteristic impedance. Kd4ttc 03:37, 27 July 2007 (UTC)
Crosstalk reduction
While "twisting of the cable reduces electrical radio interference" is correct(-ish), I don't think that the "which causes crosstalk" at the end of the sentence makes any sense. Surely crosstalk is between conductors in the cable and therefore cannot be caused by external sources (including RF interference)? —The preceding unsigned comment was added by 212.1.146.16 (talk • contribs).
- Quite so. I took a crack at correcting this. dpotter 00:29, 25 January 2007 (UTC)
- Ethernet is WELL into the radio frequencies so yes rf can be radiated by one pair and picked up by another. Afaict twisting reduces both rf emmision and rf pickup as well as some other forms of interference between conductors. Plugwash 18:49, 21 March 2007 (UTC)
Wiring diagram
![](http://upload.wikimedia.org/wikipedia/commons/thumb/5/5b/Cat5-plain-dot.svg/220px-Cat5-plain-dot.svg.png)
Is this image correct? - Zephyris Talk 15:27, 3 March 2007 (UTC)
- According to the text in the image it is a view from the bottom with the tab down, that would make it a view from the tab side (though its a very confusing way to label the image) read that way both the pin numbering and conductor order are backwards. Plugwash 18:03, 3 March 2007 (UTC)
- Which way is "up"? Is the lock-tab on top (which would make the diagram correct), or is the lock-tab considered to be on the bottom? Various workstations, servers, patch panels, switches, hubs, routers, and wall plates have them both ways, so is there a technical "right-side-up" to these or not? 10:54, 12 March 2007 (MST) —The preceding unsigned comment was added by 199.67.138.242 (talk) 17:55, 12 March 2007 (UTC).
- Most diagrams I've seen have the tab side down. -Roofus 04:24, 14 March 2007 (UTC)
- Which way is "up"? Is the lock-tab on top (which would make the diagram correct), or is the lock-tab considered to be on the bottom? Various workstations, servers, patch panels, switches, hubs, routers, and wall plates have them both ways, so is there a technical "right-side-up" to these or not? 10:54, 12 March 2007 (MST) —The preceding unsigned comment was added by 199.67.138.242 (talk) 17:55, 12 March 2007 (UTC).
It was removed from the article for being incorrect - it seems correct and useful!
- One could argue that it's incorrect because the connector shown is not an RJ-45 connector. It's an 8P8C. 128.111.207.145 18:07, 19 April 2007 (UTC)
Cat5/cat5e differences
There's no information in the article as to the difference between CAT5 and CAT5e. Just that it's "enhanced" and has "tighter specifications." 128.111.207.145 18:08, 19 April 2007 (UTC)
Turns per inch?
I was taking apart some Cat5 cable and noted that the turns per inch are different for each pair. Is that a part of the Cat% spec to improve crosstalk resistance? Kd4ttc 03:41, 27 July 2007 (UTC)
- I belive it is but I don't know where to find details. Plugwash 14:00, 4 October 2007 (UTC)
The reason for using twisted pairs is that any magnetic field impinging on the twisted pair induces offseting voltages in each twist. Other than the end effect of using only half a twist, these voltages cancel. If all the pairs were twisted evenly, the magnetic fields very likely would all line up and produce a great amount of cross talk. If the pairs are twisted at different intervals, then the magnetic fields (and corresponding induced voltages) cancel, other than for end effects.
At some interval these twists approximately repeat themselves. If you have a very bad cable and the cable is in a difficult run, but is undamaged, never kinked and is in pristine condition, one thing worth trying before scrapping the cable is to cut off a little under half of the repeat length and reconnector. This would dramatically change the crosstalk (for good or for bad) in a given piece of cable!
Twisted Pairs used
The article implies that all 4 pairs are used for both 100 & 1000. CompTIA Network+ text tells me that all 4 pairs are used only in the 1000Mbits/s environments. I think that the 100 uses only two pairs. —Preceding unsigned comment added by HarryAlffa (talk • contribs) 10:29, 31 March 2008 (UTC)
100MHz or 125MHz
If 100BaseTX is using 4B5B for a symbol rate of 125MHz, wouldn't the cable need to match this? Where is a verifiable refrence to the 100MHz? —Preceding unsigned comment added by 138.162.140.55 (talk) 21:59, 2 July 2008 (UTC)
Technical Specs
I haven't had sight of the specs so I don't know what the numbers should really be, but I think they would be useful to have on the page:
Property | Nominal Value | Tolerance | Unit |
---|---|---|---|
Characteristic impedance @ 100MHz | 100 | 15 | Ohms |
DC-Loop resistance | 170 | ? | Ohm/km |
Velocity ratio | 0.69 | ? | speed of light |
Capacitance | 300 | ? | pF/km |
Are there any objections/corrections? AndyP 13:46, 12 June 2006 (UTC)
- I'm not sure about using the source that's currently provided for the information in this table in the article. The specifications for that particular company's wire may be different than the standard(?) Correct me if I'm wrong. --66.253.230.56 (talk) 20:55, 16 October 2008 (UTC)
These values are specific to that manufacturer, and differ from other manufacturers. Can anyone provide a standards document? Who governs it -- the IEEE maybe? Pborthwick (talk) 15:28, 11 August 2009 (UTC)
- The values in the article are example values to get an approximate on characteristics. Official IEEE documents may give values that may not be reflected in real life. Electron9 (talk) 16:39, 11 August 2009 (UTC)
cat5e testing
when testing cat5 cable there are 4 tests which are done 1)wire map - which checks that all coloured cores are in the correct position. 2)length - which checks the length of all cores to ensure no brakes. 3)next and 4)attenuation - which has something to do with powerbands i think, but does anybody know what tests 3 and 4 actually mean.
- Rob 11/1/2008 —Preceding unsigned comment added by 81.159.251.151 (talk) 20:43, 11 January 2008 (UTC)
- NEXT is Near-end crosstalk. This test would measure how much of the signal intended for pin 1 is also present on pin 2, etc. Crosstalk always exists, the specification defines the limit for how much is acceptable while still providing performing full performance. Far-end crosstalk measurements can also be made.
- Attenuation is a measurement of signal strength reduction, in this case as a result of the cable itself, or connectors along the way. For example, your total cable run may only be 80 metres, but if you use 80 one-metre cables, with female-to-female connectors between them, you may have so much attenuation that the signal received at the far end is too low to be understood.
- Hope that helps your understanding. Cheers, —fudoreaper (talk) 00:14, 22 October 2009 (UTC)
Connection Diagram
Maybe this is the wrong place for it but should there not be some connection diagrams here ? Ok found it on the RJ45 Page
Individual twist lengths
Is this standardised? I have just taken apart some CAT5e and there are visibly more twists in the orange pair compared to the green.
Solid vs Stranded
There isn't any discussion in the article of solid vs stranded. Is it the case that solid is generally used for infrastructure, and stranded is generally used for patch cables?
- yeah thats what you are supposed to do. i've made patch cables with solid before though and it doesn't seem to be a problem. Never tried to terminate stranded to a wall port but i don't think it would work very well with the type of IDC the wall ports use. Plugwash 21:24, 17 August 2005 (UTC)
- Patch cables with solid wire have a risk of developing intermittent connections. Kd4ttc 03:32, 27 July 2007 (UTC)
It's important to use the proper type of connector. There is a difference between connectors for stranded and for solid wire. Much of what is sold is only for stranded wire. Using these connectors for solid wire will result in an intermittent connection. 98.204.103.241 (talk) 01:16, 29 December 2007 (UTC)
I added the following. hope it stays... its a trade secret, hard to find a reference
Solid core Cable vs Stranded Cable
Solid core cable is cheaper and is used for long permanently installed runs and care should be taken to avoid bending the conductors as the conductors may crack. The vibration from even small movements, such as from a fan, may cause cracking of solid core.
Stranded cable is used for fly leads at patch panel and for connection to movable or vibrating devices,as it resists cracking of the conductors. Stranded is more expensive and therefore commonly used for short runs.
There is a different type of crimp connector for solid core than for stranded . Use of the wrong type connector may lead to unreliable cabling.
202.92.40.202 (talk) 03:06, 7 December 2009 (UTC)
Dielectric formulas??
In the article "Characteristic impedance" telegrapher's equation states:
where
- is the conductance of the dielectric per unit length,
In the article "Electrical conductance": Conductance is related to resistance by:
for purely resistive circuits.
where:
- G is the electrical conductance (for gate),
- R is the electrical resistance,
- I is the electric current,
- V is the voltage.
How does one get the [Siemens] conductance from the values in the "Relative static permittivity" article ..?
--Electron9 (talk) 04:48, 27 January 2009 (UTC)
- You do not.
Permittivity related electric field to voltage... this is useful to work out capacitance.. the permittivity of the insulation is one of the factors to work out capacitance of the cable.
- permittivity does not relate to the loss of electric flux - its the loss of electric field flux whch is resistance. —Preceding unsigned comment added by 202.92.40.202 (talk) 03:33, 7 December 2009 (UTC)
Photo
The article says that the Cat5 standard is an unshielded cable. Correct me if I'm wrong on this, but isn't that a shielded cable in the photo? --Xcali 03:28, 15 Jun 2005 (UTC)
- I'm pretty sure you're right. The shielded connectors are certainly misleading, even if the cable itself isn't shielded, so I've replaced the picture of the (probably) shielded cable [1] with the one we had before [2]. --Elembis 00:43, Jun 18, 2005 (UTC)
- afaict most suppliers sell category 5 UTP (unshielded twisted pair) and category 5 STP/FTP (Shilded or foiled twisted pair). I don't know what the original standard says though. --Plugwash 21:19, 17 July 2005 (UTC)
- The original photo was tagged for deletion, so I replaced it with a higher resolution one with a free license. --Fo0bar 02:12, 1 September 2005 (UTC)
I find it unlikely that the wall socket on the photo CAT-5E-Wall_Outlet.jpg is actually CAT5/CAT5e compliant. RJ45 connectors compliant with the CAT5/CAT5e specification usually have a much more complex design in order to match the crosstalk requirements.Brolin (talk) 10:10, 14 January 2010 (UTC)
110m cat5
I will try to use 110 M CAT5e cable to join with a D-Link Switch for Internet sharing purpose, I expecte some slower speed becasue of the distance.I will be happy if someone can advice me how to deal with this problem. —Preceding unsigned comment added by HTINKYAWOO (talk • contribs)
- Unfortunately ethernet doesn't degrade gracefully if the cable is bad it just starts losing packets like hell (which will then have to be resent all the way over the internet). Personally in a situation like this i'd run a long ping test with large packets and see how lossy it is. Possible courses of action if it's too lossy either force it to 10MBps half duplex (don't try and force a full duplex speed on a link to an unmanged switch) or to insert an extra hub/switch in mid-run. Plugwash (talk) 01:41, 16 January 2010 (UTC)
Insulation-piercing connections
In the text, it is implied that insulation-piercing connectors and insulation-displacement connectors are opposites, but if you follow the links to either of these you are taken to the same page where it is stated that they mean the same thing. —Preceding unsigned comment added by 86.131.58.93 (talk) 09:09, 13 June 2008 (UTC)
- The answer is, they are different, they are different because there is stranded cable and solid core cable ...
- Piercing for stranded, Displacement for solid core.
- Piercing connectors are used for stranded cable. Piercing connectors would weaken the solid core, and the hole in the solid core might grow and thus cause loss of contact.
- Displacement attachment is used for solid core, as the connector wraps around the solid core, it reduces the chance of cracking the conductor. The displacement attachment might well just grip around the outside of the insulation of stranded cable, as the stranded cable will crush under the force.. when it returns to shape its width wont be as large... 202.92.40.202 (talk) 03:27, 7 December 2009 (UTC)
- I can't say I agree with you there, ribbon cable connectors use the displacement method and yet ribbon cable is always stranded. Similarlly you can get RJ style connectors designed for solid cable and they still use the peircing style (and I know from bitter experiance that RJ connectors designed for solid cable DO NOT work reliablly on stranded). It seems the difference between connectors designed for solid and those designed for stranded is pretty subtule (afaict those designed for solid have a sharp point designed to dig in to a single strand while those for stranded are less sharp and designed to crush the strands arround them against the insulation). Plugwash (talk) 01:36, 16 January 2010 (UTC)
Individual twist lengths
The reference cited [1] is purely anecdotal and this section either needs to be proven mathematically or a reference proving that individual pairs have different twists and why needs to be found. Tcatkins (talk) 15:21, 4 December 2010 (UTC)
- I've added a tag to the article and directed discussion here. Different twist rate per pair is definitely effective in reducing crosstalk. The exact numbers given here are of dubious origin. --Kvng (talk) 13:30, 5 December 2010 (UTC)
CAT5/CAT5e specifications
The article states that Category 5 has been superseded by the Category 5e specification, but treats them like they are the same exact thing. What differs in cable design, specifications between Cat5 and Cat5e and how and when did this take place? PaulCSX (talk) 21:54, 28 October 2010 (UTC)
- I agree this aspect needs elaboration from someone knowledgeable. 82.81.2.70 (talk) 00:01, 11 February 2011 (UTC)
- Yes, it needs to be more clear. Both the differences. And the title should probably be Category 5e cable instead of Category 5 cable. --Mortense (talk) 08:37, 16 April 2011 (UTC)
Cat 5 history
Question. Hi. I just wondered. There is no history here. When was CAT5 introduced? —Preceding unsigned comment added by 81.191.230.2 (talk) 08:32, 14 March 2011 (UTC)
The "Category" specification started in the late eighties to early nineties. I believe the first "Category" standard was formalized in 1991 by the TIA/EIA.
The "Category" rating system was created by the Committee as a standardized version of the "Levels" system developed by Anixter International. Anixter's intent was to provide a way for their customers to easily identify a choice in cabling to best meet the customer's application. It also gave them a way to specify a cabling standard to their suppliers; each batch of cable was tested as it was received and if it didn't meet the standard, it was sent back.
Anixter was an early proponent (if not the originator) of the concept of structured cabling; using a single cable plant for the wide variety of communication needs (i.e., serial terminal, 3270 terminal, 5250 terminal, networking (Ethernet, Token Ring, ArcNet ...),security (alarm, video), baseband video, broadband RF (cable tv, antenna)). By using "adapters" (BALUNs and impedance matching devices) the cabling system could be used for nearly every signal type. IBM also had a cabling system they were promoting, primarily for their Token-Ring networking and mainframe terminal systems.
Category 3 (the first "Cat-rated" level) grew out of IBM's "Type III" used to describe their Unshielded Twisted Pair (UTP) media.
The capacity and capability of higher Category levels evolved by changing the twist ratio (pair-to-pair), the "lay" of the pairs within the sheath, insulation material and dimension of the individual conductors, insulation and dimension of the sheath material, the size of the conductors, and other manufacturer-proprietary methods. Cat6 and above tends to have an "X member" to further isolate the pair and keep them in proper distance proportion and enhance the crosstalk figures. It takes more than pair twist to create a higher Category Rating level. 75.30.110.158 (talk) 17:41, 26 May 2011 (UTC) S. Mackenzie
Cat 5 Manufacturing
Many Cisco students are required to make a Cat 5 cable but often ask how such a complex process can be automated. Can a link be included to a manufacturers website or a video on YouTube showing how this process is automated?
Robinatilio (talk) 12:02, 12 July 2008 (UTC)
- That's an excellent request, and something i've wondered. I don't have a video, but i share your desire to see one. Cheers! —fudoreaper (talk) 00:06, 22 October 2009 (UTC)
- I always assumed that they just used cheap third world labour to make them. Plugwash (talk) 01:37, 16 January 2010 (UTC)
- Have a look here (old machines but not 3rd world I assume) --Copa017 (talk) 08:12, 1 April 2011 (UTC)
- That shows how cable is made (though dispite the title the cable they are making doesn't look like cat5 to me) but the original posters question seemed to be how cat5 patch cable assemblies are made commercially. I just don't see any reasonable way to automate the process of seperating out the cores and arranging them by color code. Plugwash (talk) 11:05, 8 September 2011 (UTC)
- Have a look here (old machines but not 3rd world I assume) --Copa017 (talk) 08:12, 1 April 2011 (UTC)
- I always assumed that they just used cheap third world labour to make them. Plugwash (talk) 01:37, 16 January 2010 (UTC)
Tautologies in lede
I'm amazed that a particular editor is still having this problem but since theat is the case maybe I need to be more explicit:
- Ethernet is a digital signal (even Gigabit)
- ATM is also a digital signal
Analog and digital are a) not capitialised and b) not signals, but broad classes of signal. In fact, if we neglect hybrid, they are the only classes of signal. Therefore we can't use them as examples of signals because they are not specific. Reducing the proposed amendments down we end up with something like "This type of cable is used in structured cabling for computer networks such as one signal and another signal, also used to carry many other signals such any class of signal."
This is undoubtedly true (neglecting the intrinsic limitations of the cable) but an obvious tautology. We don't need to confuse the reader with this, who is going to wonder what he has missed. And yes, Sgeeves, I do know what I am talking about: it is not me that is repeatedly adding a frankly idiotic statement and insisting it is some significant revelation. Crispmuncher (talk) 15:41, 12 May 2011 (UTC).
Bandwidth
The article currently claims Cat 5e UTP has a cutoff frequency of "50323 Hz" (which seems excessively precise). A cutoff frequency of about 51 kHz seems insufficient to support the 125 Mbaud used in 100BASE-TX#100BASE-TX and also in Gigabit_Ethernet#1000BASE-T. If there is some some bandwidth (signal processing) trick to getting symbol rate over 3 orders of magnitude faster than the cutoff frequency, could we mention it in this article or at least link to some other article where that trick is described? --DavidCary (talk) 17:56, 3 September 2011 (UTC)
- Examining the reference, it appears that this was calculated from resistance and inductance. In addition to excessive precision, it looks like there was an error in the calculation. I have corrected it. Below the corner frequency, the cable does not have constant characteristic impedance and thus is no longer a well-behaved transmission line. Other math comes in to play when operating at these lower frequencies. All of this either deserves elaboration in the article (and I don't feel like I can do it without original research) or some of these details should be removed. --Kvng (talk) 14:08, 5 September 2011 (UTC)
Shielding
I've just re-undone an edit made by Kvng regarding the properties of the screen. The section he cites is not relevant since the screen does not form a Faraday cage. By definition a Faraday cage is isolated to ensure an induced charge may form to counteract the applied field. In contrast, the shield here is firmly grounded (IEEE 802.3 27.5.2.2) to ensure no charge may form on the screen. By enclosing the central conductors in a screen held at ground potential it is ensured that there is no externally-applied electrical field within the body of the cable. This is why, for example, a 10BASE2 or 10BASE5 network wouldn't work reliably if the ground wire wasn't attached. It is also why STP connectors are needed - if the shield was left floating you wouldn't need to maintain its electrical continuity.
If you want further evidence of this simply take a look at the cables used by the professional audio guys - grounded shields are very much the order of the day, even though they are exclusively concerned with signals far below the 100kHz it is claimed shields are not effective at. Crispmuncher (talk) 06:30, 26 November 2011 (UTC).
- The section I cite is Electromagnetic_shielding#Magnetic_shielding. The Faraday cage discussion is a red herring. There simply is no physical mechanism for a non-magnetic shield to stop these long-wavelength magnetic fields. Bill Whitlock's tutorial (http://www.aes.org/tutorials/) explains this in depth. Since that is a couple hours long and you need an AES membership to access, I'll offer an alternate reference that simply states the facts - Jim Brown (2004). "Chield Current Induced Noise part 2" (PDF). SynAudCon. Retrieved 2011-11-27.
Nearly all interference below a few hundred kHz is magnetically coupled. Cable shields provide almost no magnetic shielding in this range.
It is important to get this right. People need to know that using shielded cable does not give them permission to bundle signal wires with AC power. Professional audio guys use shielded cable to prevent radio frequencies from contaminating their signals. --Kvng (talk) 15:07, 27 November 2011 (UTC)
Cat 5 v Cat 5e
Can we get more info on the differences between Cat 5 & 5e? I seem to remember that before Cat 6 was finalised we used Cat 5e cable for 1000Mb/s?
--Quatermass (talk) 22:57, 11 December 2011 (UTC)
Done -—Kvng 15:58, 5 January 2013 (UTC)
Signal Length
The signal length is inaccurate. — Preceding unsigned comment added by 99.119.21.148 (talk) 23:13, 4 February 2012 (UTC)
- The figures quoted here are perfectly accurate - we quote what the relevant standards say, not that some anonymous nobody noticed that a link slightly over 100m works some of the time between two particular devices. For half duplex links it's a pretty hard limit since the timings are fairly marginal (you risk late collisions) and even for full duplex links it is subject to the vagaries of the individual devices are could even vary over time. That isn't the sort of nonsense we quote here. Crispmuncher (talk) 03:11, 5 February 2012 (UTC)
Discovery channel claim Polypropylene
Discovery channels "how do the do it?" shows a Cat.5 manufacturing that uses Polypropylene. Which is not mentioned in the article. What's the truth? Electron9 (talk) 05:23, 1 February 2012 (UTC)
- The standards specify electrical performance, not construction. The article lists insulation material and their designations. -—Kvng 15:59, 5 January 2013 (UTC)
- After reading your comment Electron9, I wanted to know the specific season and episode number where you found the info, and so I found it for us all... It's: Discovery Channel, How Do They Do It, Season 1, Episode 4, "Internet Cable, Cranes, Pencils, Water, Toothpaste". Sincerely, •ː• 3ICE •ː• 12:46, 12 July 2013 (UTC)
Twist Pitch
The twist pitch table should be deleted, I don't believe there are any requirements or specification for this, other than manufacturers finding that too many twists increases length and inductance, and too few twists reduces the benefits of noise cancellation. But the actual twist pitches, and the differences between each pair is entirely up to the manufacturer, as required to meet the cross-talk and other specifications. — Preceding unsigned comment added by MitchellShnier (talk • contribs) 18:58, 21 February 2012 (UTC)
Since a copy of the ANSI spec costs $989, I'll take some pictures of 3 or 4 brands of cat5e cable with the outer jacket stripped off for empirical evidence of the de facto twist rates.Darr247 (talk) 21:46, 31 December 2012 (UTC)
The twist rates and colors are specified in section 3.4 of ANSI/ICEA S-90-661, which IS a referenced document in the TIA-568 testing specification, and even though that one's only $116, I doubt they would consent to posting that section of it here. I would enter that link as a reference in the wiki, but would prefer not to delete your 'dubious - discuss' citation. i.e. feel free to do so. Darr247 (talk) 04:49, 3 January 2013 (UTC)
A reliable source for the information in the table is desperately required. Although the technique of varying the twist pich is well known, a source for the actual pitches is required. The original supporting reference has had to be deleted as it admitted that the twist pitch was measured and thus was original research. In any case it is an enthusiast's site and thus not a reliable source. 86.166.70.75 (talk) 19:13, 22 July 2013 (UTC)
- User:Darr247 claims twist pitch is specified in ANSI/TIA/EIA-568-A and ANSI/ICEA S-90-661. We just need to find an editor with access to these standards to confirm. Or User:MitchellShnier is right and it's special sauce formulated by cable manufacturers. ~KvnG 12:45, 25 July 2013 (UTC)
HDBaseT
Hi there, AV expert and first time wikipedia editor :-)
There are two problems with the HDBaseT statement in the Applications section. 1) The link to the reference to the HDBaseT press release no longer works. The HDBaseT alliance must have taken down or moved the press release. 1b) There should probably be a more specific article or source discussing the use of UTP for HDBaseT than a press release. 2) The statement that Cat 5 may be used for HDBaseT is conditional and should be expanded as to not imply that any Cat5 cable will deliver full HDBaseT functionality. Obviously a Cat5 or Cat5e cable rated at 100MHz will not transmit 10.2GBps even when using all four pairs. There should be a comment that higher-resolution video formats require greater bandwidth on the UTP cable (such as 350MHz or 550MHz), and many manufacturers of HDBaseT endpoints recommend Cat 6 or Cat 6A cable for full performance.
I know it's just an example of an application, but this just bothers me a but. Anyone mind if I go ahead and fix it? --Kevinzolitor (talk) 15:47, 7 August 2013 (UTC)
- Welcome to Wikipedia! I would prefer to just remove the HDBaseT mention. HDBaseT does say (with references) that either Cat 5e or Cat 6 will work. Maximum bit rate in bps is dependent on channel capacity can be much higher than bandwidth in Hertz. ~KvnG 19:51, 7 August 2013 (UTC)
applications - multiple signal types in the same cable
Under Applications it is stated without Citation that it is ok to run up to 2 phone lines along with an Ethernet run (use those "extra" pairs, why not?). This is all well and good until somehow the phone decides to ring and you now have a 100V+ ringing signal across the primary of what amounts to a 100M long transformer. In the best case this ringing signal raises the noise on the other pairs, often to the point of interrupting the data link. In the worst case, after such repeated abuse, the switches and Ethernet cards on either end tend to take exception and quit working completely. please show me a citation showing that this is a non issue. — Preceding unsigned comment added by 71.105.114.179 (talk) 07:42, 23 August 2013 (UTC)
- I added a few citations, as you requested, to Category 5 cable#Applications supporting "2 analog phone lines and Ethernet on a single Cat5 cable".
- Because of the way twisted pairs work in a Cat5 cable, very little of any signal (including the ring signal) on one twisted pair couples to any other twisted pair.
- Even if the ringing signal did completely corrupts and interrupts the data link during each ring -- so what? The Ethernet protocol is designed to re-send any packets that are corrupted or interrupted.
- --DavidCary (talk) 18:31, 28 April 2014 (UTC)
CAT 5 VS 5E
Seems to me these are not compatible. Reference? Seems important.
- This is explained (with reference) in the third paragraph of the lead. More detail and additional references are in Category_5_cable#Category_5_vs._5e. ~KvnG 13:30, 3 June 2014 (UTC)
True gigabit
84.239.241.254 asserts that Cat 5 can't support "true gigabit". It is true that Cat 5 was deprecated and replaced with Cat 5e around the same time that Gigabit Ethernet came out. This was done to correct the omission of a crosstalk specification in the Cat 5 specification. Other than using all 4 pairs, Gigabit Ethernet is no more demanding on cables than Fast Ethernet. Since Cat 5 is depricated, Cat 5e is required for Fast and Gigabit Ethernet. Cat 6 is not required for Gigabit. If you believe I am mistaken about this, please explain and please specify where in the thousands of pages of IEEE 802.3 you find this. ~Kvng (talk) 15:33, 6 May 2015 (UTC)
Individual twist lengths
This section is marked as "dubious," but any freshman physics student can tell you that the characteristic impedance of a parallel-conductor transmission line (TL) depends on only two mechanical parameters: the diameter of the conductors and the spacing between them. Such cables are twisted in cases when possible for two reasons. First, assuming uniform insulation thickness, this ensures that the spacing between the conductors is constant and as tight as possible (eliminating "slack" which would allow for crosstalk or RFI). Second, it "folds" the enclosed area over and over so that any local magnetic flux will destructively interfere. I'm sure that as the pitch approaches the diameter or spacing of the conductors the underlying physical assumption of truly parallel conductors becomes less relevant, but in the case of cat5 cable the length of a single twist is many times greater than either. In other words, viewed from the perspective of an electron, the two conductors are effectively 100% parallel. When you look at a 300-ohm TV feedline you can see that twisting some parallel TL's is impractical, as it cannot be done without significantly altering the spacing between conductors.
It is advantageous to twist multiple pairs in a single cable at different pitches to further reduce crosstalk, because if all pairs were twisted at exactly the same rate, there would always be a fixed geometric relationship between any one conductor and its seven neighbors; in this case it would be as if there was no twist at all with respect to crosstalk. Varying the twist keeps any single conductor's influence from compounding over the length of the cable; comparing two points along the cable, flux between pairs A and B is just as likely to interfere destructively as it is constructively. You can demonstrate this by solving a couple easy integrals.
There are literally thousands of web pages out there that explain this in great detail. Simply searching the web for "characteristic impedance of parallel feedline" will get you all the references you need, but it could be argued that this is such a fundamental piece of information that no reference is needed. You don't have to explain that "1+1=2" every time you make an arithmetic assertion. I recommend that we first identify exactly what piece of information is held in dispute, because in my professional field, every bit of this paragraph is implicitly accepted as cold, hard fact, thousands of times a day. In fact, it's something we depend on. Hope this helps! 192.107.155.5 (talk) 18:52, 8 May 2015 (UTC)
- There is discussion on this at Talk:Category_5_cable/Archive_1#Twist_Pitch. It looks like the concern there was not about impedance but whether twist pitch is specified in the standard. It think it was resolved that pitch is not specified. We probably should just remove the {{dubious}} tag. ~Kvng (talk) 19:05, 8 May 2015 (UTC)
Category 5 vs. 5e
I've trimmed this section back to what's actually said in reliable sources. The Cat 5 specification has been deprecated. Cat 5e is required for Fast and Gigabit Ethernet.
Cat 5 was underspecified and so, in theory, it was possible to build a cable that met specifications but did not work for Gigabit Ethernet or full duplex Fast Ethernet. Yes, you're more likely to notice this problem with high data rate transfers because the problem will corrupt one out of n bits (where n is quite large) so the more you transfer, the more errors you see. ~Kvng (talk) 17:06, 1 June 2015 (UTC)
- The way it currently looks makes it appear that Cat5 was just fine to run gigabit but that it's not tested for it. That's really not true at all, cat5e requires a specific twists per pair and twists per foot of each pair. Cat5 was the wild west because these twists were not in the spec. Gtwy (talk) 07:34, 5 March 2016 (UTC)
- None of the category cable standards specify details of how to build a cable. The standards specify performance requirements. ~Kvng (talk) 02:09, 9 March 2016 (UTC)
- ^ "Transmission Line Zo". 090113 prc68.com