12 ms·
Is Your Cat 6 Cable a Dog? (2013)
- ramenmeal 8y agoAnyone willing to suggest a cable that meets cat 6?
- tareqak 8y agoBlue Jeans Cable (they wrote the article): https://www.amazon.com/Certified-Blue-Cable-Assembled-Jeans/dp/B00IITTEG6 https://www.amazon.com/Certified-Blue-Cable-Assembled-Jeans/... Update: This link is to their Amazon store page: https://www.amazon.com/stores/node/9182265011 https://www.amazon.com/stores/node/9182265011 .
- softfalcon 8y agoI bought 2 of these at 75 feet for $16 CAD (each) and it's maintaining expected 10 Gigabit on my home network. https://www.amazon.ca/gp/product/B00H41LYGM/ https://www.amazon.ca/gp/product/B00H41LYGM/
- Dreami 8y agoDoes "maintaining 10 GBit" only confirm Cat6? I don't want to * on your recommendation, just want to be nit-picky about these things after such an article :)
- earlz 8y agoI would think the real test is to do some pulling on the cable to make sure it's reasonably sound structurally, and also maintaining 10 GBit at the spec maximum for cat-6... but really if it works for however long you run it, what does it matter
- nomel 8y agoI used to work work for a company involved in 10GE some years ago. I believe maintaining would be no link drops, but would include fast-retrains, where the cable coefficients are quickly synced without losing link, resulting in some number of dropped packets around the event. Low signal-to-noise will result in a higher bit error rate and sporadically dropped packets. Adding aggressors (noisy fluorescent tubes, power lines, RF transmitters, etc) would eat away at the margin. But, if you're doing short runs (like 30m), then you've got absolute loads of headroom, at least with the fairly standard equipment we were testing. And also, the bit error rate is for full speed. If you don't have constant 10G traffic, then your error rate will be proportionally lower since there will be a bunch of "idle" time. When testing these, there's a bit error rate test performed at different lengths of cable. IEEE used to recommend something like 10^-13 (not sure if this has changed), but we would test, with some margin, at the full 100m using cables wrapped on spools (no external aggressors). We would also test with a "5 around 1" configuration, with 5 other fully active cables tied around the cable under test. If I recall correctly, this test required the full 10^-13. All of these cables were periodically verified to meet all specs. But regarding cable quality, we would buy thousands of meters of cable without issues, from good vendors. You may have trouble finding compliant cable from China. If there was a problem, it was usually from someone bending and putting a permanent kink in the cable (causing a discontinuity in the impedance) or closing a door on it.
- softfalcon 8y agoNo worries. Be as picky as you want to be. I merely recommended the only cable I can actually recommend for the average consumer. Was sure to add for my home office for some context. This cable fits the bill for running a connection between access points in my home. The speed is top notch, not seeing drops in speed for the many times I've moved large files across the network. Cabling seems sturdy and rugged, even while I was feeding it through various locations. As long as it appears reliable and I get the speeds I expect, I don't really care much beyond that. It works, it was cheap, I don't ask many more questions. Would this stuff hold up in a data center? Probably not. When I was wiring up homes or buildings, I always bought a high quality spool of Cat 6 from a trusted local provider. They asked for a recommendation, I did a mental "what are they likely asking? As someone in their home, or as a data center IT worker hooking up racks of servers?" I work remotely now, so I'm at home. Just assumed they were at home or similar. Maybe not the answer they were looking for. shrugs
- CamperBob2 8y agoSeems like the actual issue is that the specs are way too tight. If 80% of the cables fail the test but virtually all of them work fine in the real world, what does that say about the test? How common is it for someone to buy one of the retail-grade cables they're discussing in this article, and find that it doesn't work reliably?
- phinnaeus 8y agoMore specifically, for a digital signal, what does "working, but not up to spec" look like? Occasional packet loss? Corrupted data that is masked by ECC?
- sigstoat 8y ago"working, but not up to spec" can mean "works today, stops working tomorrow when subjected to a bit more nearby electrical noise". 10GbaseT links frequently can't negotiate down, so your link can just stop working whenever.
- russdill 8y agoAnything "masked" by ECC is part of the design.
- sannee 8y agoCan also mean: Works with devices A and B, stops working with devices A and C.
- secabeen 8y agoIt also is a testament to the ability of the interface vendors to get solid connections of out-of-spec cables. I have a small server room that is only wired with a 200+ foot run of in-wall Cat5e cables that the building was originally wired with. Running fiber to the room would be a multi-thousand-dollar job. On a lark, I took two switches with 10G-BASET interfaces and successfully brought up a 10Gigabit link across that Cat5e. I brought that link into production 6 months ago, and through that time have had less than 100 errors per link. Just incredible.
- gok 8y agoI wonder if Cat 7 cables are similarly bad. I would have thought that since you can visually verify that the pairs are all shielded, it would be harder to fake it.
- Dylan16807 8y agoCat7 is in a weird spot. It's been rendered basically obsolete by 6a, which has performance that's almost as good and allows for full-distance 10gig. And while 7a was specified with the goal of supporting 40gig, it looks like 40gig is going to require cat8. And you're not really supposed to put RJ45 plugs on cat7/7a either. So while you can verify the foil is in there, it doesn't say all that much toward the total quality, since they're not quite a standard cable at all.
- Abishek_Muthian 8y agoBut isn't Cat7 better at shielding compared to Cat6a?
- Dylan16807 8y agoIt has stricter requirements for the most part. But whether you're better off buying Cat7 with nonstandard plugs, vs. a model of Cat6a where they used shielded wires? I'm not sure.
- wrs 8y agoI've been buying Blue Jeans cables for a long time (starting with audio/video cables, and more recently Cat6) because I love their no-b.s. attitude. And if you're in Seattle and ask nicely they may even let you drive over and use their fancy tester on your existing cables. On the A/V side their story was basically "this is the same cable that the studio that made the content you're watching is wired with, and you don't need anything more than that".
- Mathnerd314 8y agoI guess the article isn't dated but judging from the Internet Archive it should have a [2013] tag. http://web.archive.org/web/20130727071218/http://www.bluejeanscable.com:80/articles/is-your-cat6-a-dog.htm http://web.archive.org/web/20130727071218/http://www.bluejea...
- handedness 8y agoThanks–updated.
- StillBored 8y agoFirst I haven't a clue why anyone would run cat 6, either do good 5E or standard 6A. 6 is well not really useful.. All that said, the article does manage to skim over a few things it should have talked about. There are actual physical differences between the cabling specs which are a result of the 6A signalling requirements. The termination is _MUCH_ stricter and likely the cause of 99% of the failures in cases where the cables actually are 23AWG, properly twisted and isolated. Its these latter three things which will allow you to identify "real" cat6A rather than the stuff being relabeled by cable manufactures claiming to have found some secret method of making what are basically cat5E cables 6A. (back to my 5e comment. The thing to keep in mind about a lot of this is that the standards specify minimums. The whole cat5E grouping was one where there was a race to out-specify everyone else, which is why you can find cat 5E 500Mhz labeled cabling, although most places just renamed it to cat 6 and charged a few cents more per foot. The other thing to keep in mind is that these specifications end up being given in loss/crosstalk per foot, so its quite possible to run higher data speeds on short cables that are non compliant per foot, particularly if they are terminated properly. Hence other comments I've made on this site about running 10GbaseT in my house on cat5E without any apparent signally/packet loss due to the runs all being very short.).
- forbiddenlake 8y agoWhy is 6 not useful and why is 5E/6A better?
- ixf 8y ago6 is useful for non-dense high-bandwidth environments where you might need to get 10G-BASET a little further than 5e will manage, but 6 doesn't address enhanced alien crosstalk. This means in dense environments it may struggle. 6A does, via shielding, and so will work better in dense environments (where you're likely to see 10G-BT anyway).
- Dylan16807 8y agoShielding's optional for 6A so be careful where you buy if you want that.
- brink 8y agoI just bought 1k feet of Cat 6 cable and ran it through my house two days ago. Would have been useful to know then.
- BuildTheRobots 8y agoCat-6, pah. Use barbed wire like a real sysop... https://news.ycombinator.com/item?id=15910263 https://news.ycombinator.com/item?id=15910263
- M2Ys4U 8y agoOr wet string: https://www.bbc.co.uk/news/technology-42338067 https://www.bbc.co.uk/news/technology-42338067 / https://www.revk.uk/2017/12/its-official-adsl-works-over-wet-string.html https://www.revk.uk/2017/12/its-official-adsl-works-over-wet...
- tossandturn 8y agoOn the other end of the spectrum (not testing specifications, but just simply "does it network?"), I have successfully ran a 100BASE-TX connection between industrial equipment using individual (not twisted) 24-AWG wires over 20ft.
- ocdtrekkie 8y agoSeeing what people have done with legacy phone wire sometimes is just disturbing.
- davidgerard 8y agonow having flashbacks to dialup tech support in the late 1990s. Thankfully I've lost the superpower of telling precisely what's wrong with your phone line and where and how many excess devices you've plugged into it just by listening to a connection noise.
- cr0sh 8y agoWhen I was in high school, I helped my friend get cable TV back to his bedroom by patching it over a really long length of phone cord. Crappy picture, but it was better than nothing, and he didn't have to share the living room TV any longer (his older brother later ran proper coax).
- llama052 8y agoI'd argue that if you're not doing a home setup, you should use fiber for anything 10GBe+. Less latency and power draw.
- meitham 8y agoCan Fibre support PoE? As that's really useful at home.
- bjoli 8y agoI am curious what you are doing that is 10GB+ on your home network. Not snark. Genuinely curious.
- meitham 8y agoI'm not using 10GB! I was not implying that. I am however using cat6 cables and I get around 800mbps speed around the house. I have PoE injector and some of my devices (access point, cameras, raspberrypies) are powered through the cable. I was genuinely wondering if one of the disadvantages of optical cables is the lack of PoE, due to the lack of copper wires.
- sokoloff 8y agoFiber is often chosen specifically for its electrical isolation properties (runs between different buildings as just one example). Your intuition that fiber doesn't support PoE is correct.
- LeonM 8y agoI don't know if you are being sarcastic or not, but there are optical cables that are bundled with copper conductors for power delivery.
- meitham 8y agoNot sarcastic at all, I was genuinely wondering, and thank you for pointing out the copper wire in optical cables, I didn't there was such a thing.
- Havoc 8y agoThey kinda gloss over one point - the ones that passed/semi-passed are the short ones. For a 1m patch cable noise rejection, cross talk & similar length isn't really an issue. Like it or not manufacturers get away with it because in the vast majority of situations a cat5 or even a failing cat5 has no real world consequence. It fails the standard yet still works in practice - on a physical level. Plus the network stack built on top of that has further fault tolerance. If it's a long cable at the edge of tolerance or you're in a mission critical datacenter then yeah maybe worry about this stuff. Everyone else just grab a patch cable off amazon.
- kiallmacinnes 8y agoThe specs define the ratings in terms of length though, don't they? That suggests to me the tester takes the cable length as an input, and gives a valid rating for the length?
- amluto 8y agoI would love to see an Ethernet spec for >100m runs, perhaps at lower speeds. If I could do 1 Gbps at 150m on Cat 6A cable without egregiously exceeding the spec, it would be great.
- Dylan16807 8y agoAs long as you meet the signal integrity standards, I don't think you majorly violate anything. So run one speed level slower than your cable normally supports and do some math on crosstalk per meter.
- rocqua 8y agoNot sure whether this still holds for 1Gb and faster ethernet, but there used to be a maximum length due to collision detection. An ethernet frame has a minimal length, and since the cable (used to be) a shared medium, frames could collide. Such a collision could be detected by senders whilst sending the frame, allowing for a re-send. However, when the cable is long enough, you could get collisions happen after you are finished sending. I'm guessing that the newer standards that only allow point to point connections invalidate those requirements though.
- Dylan16807 8y agoThat's not actually true. When the 100 meter limit was established, a minimum-length frame would be 10 kilometers long.
- amluto 8y agoI'd be more worried about skew. CAT5e and 6a seem to have the same delay skew spec, and I can easily imagine Ethernet PHYs that are simply unable to operate beyond 100Mbps with out-of-spec skew. CAT7 seems dubious (not recognized at all by everyone in the industry) but has a tighter delay skew spec, so this would help. Another potential issue would be if the PHY has an equalizer that can handle reflections due to not-quite-perfect termination at 100m or closer but has trouble beyond 100m. Something like a 1000BASE-T-LR or something that just means "1 Gbps over CAT [whatever is needed] if properly installed at lengths up to 200m" would be quite nice. Most likely a lot of equipment would meet that spec out-of-the-box.
- sinuhe69 8y agoFor our projects, we always sourced directly from big names in the industry like Nexans and LeGrand. With a hefty price, of course. But I always wondered whether their cables are worth the price. Of course, we have let our system integrators tested and reported diligently but yeah for big money, rigging is not so far fetched!
- slothtopia 8y agoI disagree I prefer cat 6's design more streamlined so the photons travel more aerodynamically
- otterpro 8y agoIs there any testing equipment that can test for compliance, and is inexpensive? Fluke cable tester (DTX-1800) that was mentioned cost $40,000! I saw an eBay listing for $6500 for used one. There are other cheap testers out there but they seem to be just continuity tester, and doesn't test for cross-talk, etc.
- Already__Taken 8y agoThe cheapest test equipment is going to be some spare computer hardware at each end and a passing a test-suite of your expected performance numbers.
- therealx 8y agoThe IDEAL LanTEK III-500 is 5kish.
- vinceguidry 8y agoI'm going to be changing out the flooring of my condo at some point and am wondering if there's a decent solution for underfloor wiring that doesn't involve building a raised subfloor. Or perhaps I could build a very small one, I just wouldn't want to just wing it.
- gog 8y agoI did just that few months ago. I had an electrician come and make channel in the floor, lay in it some plastic tubing and later he pulled the cables through it.
- jabl 8y agoSemi-off-topic Q: Why did twisted pair cables become the standard for ethernet? Seems to me from a signal integrity standpoint that coaxial (yes, there was "thin" and "thick" ethernet back in the day) or twinax type cables would be much better? Were they originally cheaper than coax, or what is the reason?
- eyeareque 8y agoBriefly, twisted pairs reduce noice and crosstalk.
- alanfranzoni 8y agoIIRC coaxial was terrible back then. It was stiffer and hence harder to place around corners, and the connectors were harder to clip on. I don't know if those reasons were important or even considered, back then.
- avian 8y agoCost and simplicity of installation. Twisted pairs and crimp-on RJ connectors were already widespread in telecom due to POTS. If not for anything else, they were cheap and their reliability well understood due to economies of scale. On the other hand, coax cable was significantly more expensive per length. Compared to RJ, BNC coax connectors were expensive and hard to correctly crimp onto the cable. When computer industry tried to make them cheaper it also resulted in very crappy products. Unreliable connections were cause of most of the gripe with older coax-based networks. Which is kind of counter intuitive. A properly made BNC is very reliable and still used extensively in electronics test equipment and RF applications. But apparently reliability just didn't scale down cost-wise.
- flyinghamster 8y agoI also don't miss the daisy-chain wiring of coax Ethernet. Any one connector getting out of whack, or a bad/missing terminator on either end, would take the whole network down.
- 8y ago
- eyeareque 8y agoOne thing to watch out for is CCA cat cabling. It’s cheaper, because it’s copper coated aluminum. I have no clue if it reaches spec, but i don’t think it’s worth it because it would cost so much to rewire your house again later (or trace down an outage/packet loss in production).
- kokx 8y agoI actually have CCA cabling. On purpose. Why? It's cheaper. And I only had to run about ~15 meters through my small rental apartment. And it basically cut my cabling cost in half while having enough reserve cables left. It hasn't failed on me so far. When I buy my own house, I will definitely switch to better cables. But in such a small rental apartment, which I won't be living in too long, I didn't find it worth my money to buy better cabling.