Cat 6 cable failure

0

I laid a direct buriable what I thought was high quality Cat 6 cable over about 200 feet, worked fine for a couple weeks then suddenly failed. When I attach the tester to each end, the lights barely light up dimly except for #5 which flashes bright. I have attached new connectors to each end twice now, four gold plug RJ45 wasted, same exact result.

I tested the two routers being connected with a 30' ethernet cable, worked fine, also tested the tester itself with this 30' cable, it flashes all 8 digits brightly and correctly.

Have examined the length of the cable there is no damage no cuts in it, and it's only a couple weeks old anyway. All I can think of is that there are a few bends around corners including a couple that are almost 90 degrees, could this over a two week period lead to failure? Bizarre.

CrYellow

Posted 2018-12-17T18:17:55.933

Reputation: 11

2A sharp bend is fatal for CAT5 / 5e / 6 / Coax cable... You need to be gentle with it! Can you provide photos? – Attie – 2018-12-17T18:25:01.677

The very best case is that crosstalk between the pairs becomes too high and the link fails, the worst case is that the conductor actually breaks (which is what sounds like happened for you). – Attie – 2018-12-17T18:25:54.540

4I don't know if this is the cause in your case, but a too-tight turn can break internal wires in an Ethernet cable. The fact that it's 90 degrees is less important than the "turning radius" or "bend radius", which the manufacturer should have published. If the cable moves or vibrates at all or is under tension, the internal wires can easily break. – Christopher Hostage – 2018-12-17T18:26:02.153

@CrYellow these comments are wrong. 90 degree or greater bends will not damage your cable. See my answer below. – Keltari – 2018-12-17T18:37:41.633

1@Keltari I don't know where you get that info, but obviously a 90+ degree bend goes beyond the minimum bend radius of Cat6. When a cable is bent beyond it's minimum bend radius, it can cause transmission failure, damage it, or shorten it's life. The minimum bend radius of Cat5e and Cat6 is about 1 inch. A few possibilities of the problem would be a kinked line, or the cable is stapled too hard. Tip: use velcro instead of staples. – DrZoo – 2018-12-17T18:53:40.113

How did you terminate the cable (punch down to sockets or connectors directly on the wire)? Did you have the correct test equipment (very expensive) that can properly run the cable through the category test suite? Most cable failures are due to installation. Even experienced installers can have problems with Category-6 and Category-6a. Exceeding the minimum bend radius (four times the cable diameter) or maximum pulling tension (25 lbs.) can permanently ruin the cable.

– Ron Maupin – 2018-12-17T18:57:07.197

@DrZoo honestly, common knowledge. There is a great post on this site doing a physical Ethernet fable torture test. 90 & 180 degree turns, wrapping around a nail, being stapled, etc. The car worked fine. Additionally the OP said the cable has continuity on all wires. The problem is resistance. – Keltari – 2018-12-17T19:40:51.100

I don't see how bending would matter because if bending were an issue cable would have failed right off the bat, this failure after two weeks means that the cable somehow went bad. As Keltari knows further testing shows connectivity between only the two blue wire sets, there is no connectivity between the orange green brown wires when twisted together like this https://i.imgur.com/brxyCpU.jpg at one end, and tested for continuity at the other end with multitester.

– CrYellow – 2018-12-17T20:09:18.270

@Keltari Again, going past the bend radius does not guarantee that you will not damage the cable or shorten it's lifespan. Torturing your cable by wrapping it around a nail, stapling it too hard, kinking it, etc is only increasing your chances of damaging the wires inside. Using a bit of cable care and common sense, will help reduce the chance that you'll have to make the run again. – DrZoo – 2018-12-17T20:10:57.390

@Keltari and everyone else I started pulling the cable and found that it had been chewed through at some point where laid flat on the ground whatever chewed through the outer split conduit and the cable itself apparently not enough to sever entirely but enough to cut all but the two blue wires, as I was pulling the cable it separately entirely. Wasn't separated before which is why I couldn't see the damage. https://i.imgur.com/YQo2YVe.jpg https://i.imgur.com/nfFJMvE.jpg https://i.imgur.com/HQR5gwB.jpg No joy in Mudville must think of a different route or conduit are hungry critters here.

– CrYellow – 2018-12-17T20:37:28.690

Anyway, in a roundabout way this "proves" that bending is not an issue, but a chewed through cable definitely is! – CrYellow – 2018-12-17T20:52:53.860

Apparently, you misled everyone in your question: "Have examined the length of the cable there is no damage no cuts in it..." The comments are accurate in the context of the question. – Ron Maupin – 2018-12-17T21:23:54.343

I didn't mislead anyone, I did not observe anything at that time. You're a bit of a kook saying something like that. Obviously - you think I would have been spending all time time furiously conducting tests if I knew the cable had been eaten? --- And anyway, now the cable is reconnected and working fine, BENDS and all. So obviously bends are not an issue whatsoever. – CrYellow – 2018-12-18T19:17:16.957

I did not observe anything at that time and had spent time that night with a flashlight and again that morning observing the length of the cable, just didn't notice it yet because wasn't severed entirely hadn't come out of the split conduit. When I went to pull the cable, after conducting all these tests, it severed as I was pulling it and then I saw the split. "Mislead" implies intentional action, lol. Not the case here! If anything, I misled myself into a lot of unnecessary work. Still, BENDS are okay, @Keltari was right. – CrYellow – 2018-12-18T19:25:33.857

Answers

0

No, 90 degree or greater bends will not cause a cable to fail. And the fact you get a continuity in your tester proves it (as opposed to what the comments say in your question).

More than likely, you were not using an outdoor rated cable. Outdoor rated cables are waterproof. You can use non-outdoor rated cables, if they are inside a waterproof shaft. Water will penetrate the non-outdoor rated cables and corrode the copper. This corrosion is hastened by electricity flowing through the copper. The corrosion increases the electrical resistance, which causes the dim lights on the connector.

As a side note, gold connectors are pointless. A chain is as strong as its weakest link. Therefore, unless the entire cable was made of gold, the gold connectors provide no benefit.

Keltari

Posted 2018-12-17T18:17:55.933

Reputation: 57 019

@CrYellow "supposed to be" is not actually. Use a multimeter and test the resistances in each of the wires. I strongly suspect they have shot up. You said one lit up brightly, use that is the control. – Keltari – 2018-12-17T18:41:30.237

This is an image of the tester as they sequence most of the lights light up very dimly except for #5 flashes bright. https://i.imgur.com/mblKrht.jpg https://i.imgur.com/STi5xA0.jpg

– CrYellow – 2018-12-17T18:44:09.667

I have this Digital multitester https://i.imgur.com/zSRSZUl.jpg but I'm not sure how to use it over a 200' long piece of cable.

– CrYellow – 2018-12-17T18:49:23.320

@CrYellow the fact the lights light up proves continuity. The reason the lights are dim are due to electrical resistance. If the cable is still buried, you cant test it. However, this is absolutely the reason. – Keltari – 2018-12-17T18:51:17.500

whoever downvoted, feel free to actually explain your reason, then I can prove you wrong. – Keltari – 2018-12-17T18:53:10.640

1@CrYellow You can just connect all the wires together at one end and measure from the other end. You're looking for differences between resistance measured between various pairs of pins. – David Schwartz – 2018-12-17T18:53:22.403

Okay here's what I did, I connected (twisted) the two greens together, the two oranges together , the two blues together, the two browns together. Using the beeping connectivity test, only the two blue wires even registered a beep - the other sets of wires aren't even connecting apparently. Then when I tested the resistance on the set of blue wires, it wasn't bad, although it fluctuated wildly was around 4 ohms. Tried again like this https://i.imgur.com/brxyCpU.jpg and got the same results. So I'm apparently not even getting connectivity on three out of four of the sets of wires.

– CrYellow – 2018-12-17T19:32:58.310

@CrYellow you said all wires lit up with the cable tester. You have continuity. The resistance might be so high the tester isnt registering. Test with the other tester again. – Keltari – 2018-12-17T19:43:22.067

@Keltari here is the video for download https://www.dropbox.com/s/t806r7l7lo092ny/IMG_6774.MOV?dl=0 basically 4,5,6,7 flash together very dimly for while, then suddenly 5 lights up brightly. There was a very brief dim flash at 3 too. I used T568B order for the wires, so that seems to make sense that the #5 lights up, which would be among the blue, no?

– CrYellow – 2018-12-17T19:52:40.243

@CrYellow it's hard to tell with the tester is doing. However, the cable did work. Now it doesnt. You tested the switches, they work. There really isn't anything else it could be other than cable failure or connector failure. If you are confident you did the connectors properly, that leaves the cable. To be 100% you will have to pull the cable and test with both ends accessible. – Keltari – 2018-12-17T20:01:51.490

@Keltari yes I tested the cable with the tester immediately after failure, then cut it added two different connectors with boots, result same exact tester dim lights plus bright #5, then tried again with new gold connectors at each end, same exact tester result...and now this multitester connectivity test shows that no connectivity between pairs except the blue pairs I think that's conclusive that it is the cable not connectors, no? Plus the fact that it worked for two weeks then suddenly failed. – CrYellow – 2018-12-17T20:06:45.117