Is your CAT6 cable a dog?

timeshifter

Well-Known Member
Reaction score
2,388
Location
USA
In my ongoing saga of network wiring infrastructure projects I was dong a Google search for "how to certify CAT6 cable". I ran across this article.

https://www.bluejeanscable.com/articles/is-your-cat6-a-dog.htm

Basically they claim that 80% of CAT6 patch cords they bought in store and online from name brand vendors do not pass test for CAT6. Some even failed CAT5e.

So, how do I tell if the installed infrastructure is good?

I'm working on two projects right now.

One is a small office with about 8 people. They've taken over some neighboring space and bumped out their space. I had a contractor install 14 new drops, and having some real concerns with his install. See https://www.technibble.com/forums/threads/wiring-contractor-installs-used-patch-panel.82667/

The other is a small industrial type company with about 30 employees now. They bought a building two years ago that had a bunch of amateur installed CAT6 wire that's a mess, and they're using it. See https://www.technibble.com/forums/threads/how-would-you-clean-up-this-mess.82453/

I have borrowed a JDSU ValidatorPRO NT1155 from a friend. I think he said this will certify the cabling. I played around with it. Basically it sends a file to a remote unit and gives me a pass or fail. Is that certifying? I know it can run other tests and save the results for an entire job, but I was just experimenting with the manual tests.

BUT, I don't think this tester tells me anything about my CAT6 wiring. It only transmits at 1 Gbps. Wouldn't a CAT6 test run at 10 Gbps?
 
Yes, a proper test of CAT6 cable attempts 10gbit, you'll only get a majority percentage of it, but you'll get it.

Testing at gigabit is silly, that's not even CAT5e testing, as 5e can go a bit beyond gigabit.

But yes, you're right the tools often do not care about the extremes, CAT6 is sold for guaranteed gigabit capability, so that's all anything tests for.
 
It's not that they are not proper it has to do with real world vs on paper.

If you really want to spend money Hubbell will sell you certified cat6 patch cables.

Used them on a job a month ago nice cable but extremely expensive. Customer wanted everything cat6 certified end to end including patch patch cables.

Rather than attempt to certify over 400 patch cables we opted to go with ones certified from the factory.

Sent from my SM-G870W using Tapatalk
 
That article's just another FUD piece in my opinion. For starters there's nothing to indicate that they had a proper statistical sampling plan. Don't even mention the number of cables tested from each vendor. Another thing I find very interesting is that, for the half dozen fail reports I looked at, it was the same category each time. That in of itself makes me question things. Not calling them liars mind you. More like something is not authentic because the label is in the wrong place. Of course Fluke's going to spread FUD as well to sell more equipment.

This piece by Fluke does a good job of explaining the differences in cable tests. I've got a NT1155 as well and it falls in the category of a qualifier, but no Cat6. I've found it more than adequate for my needs. The only time you need a real certifier is if you have a contract that specifies it or you want to take advantage of a cable OEM's life time guarantee.

Personally I don't worry as long as the network tests to needed customer requirements. So 1gig is more than needed in 99.999% of the cases. As far as what I install it's all Cat6 these days. Price is almost the same as Cat5e.
 
@timeshifter Yep, real world vs paper.

But there's another subtext here... mislabeled junk.

I cannot tell you how many boxes of "cat6" I've returned because it's obviously NOT CAT6. No isolator cross in the center, undersized conductor, obviously mismatched twist rates between pairs.

I'm all but certain there are some manufacturers out there that just took their CAT5E production lines, changed the labels, and moved on. And that's not OK! Dealing with it isn't terribly easy though, and testers that can actually fully test CAT6 to specified 10gbit speeds are rare, and expensive. So most of the time, everyone does what @Markverhyden does. I can't say it's wrong, but it doesn't really feel right to me either.

Gigabit may be OK for today, but with gigabit internet connections less than a decade away, they won't be for long.
 
@timeshifter No isolator cross in the center.

The spline is not a mandatory part of the Cat6 spec from what I've read. When I first started buying Cat6 several years ago all of them had an X spline. In the last year or so I've started seeing Cat6 with a flat spline or no spline are all. Now, mind you, this is Cat6 not Cat6a. The characteristics of construction specified by the standard is really limited in scope. The big thing is the wire gauge is 22-24. I suppose they could use corn silk as long as it met all the test specs.

What I don't like is there is nothing even close to economical for a Cat6 tester. I'd consider 1500-2000 economical with all the right features. A lot advertise Cat6 but that's only because they report a fail if there are not 4tp's, which is a requirement for Cat6.
 
Getting bad patch cable is an ongoing issue and has been written about many times.

Here are a few articles for reference. Some of these are old articles, but it shows that the problem has been around for a long time.

https://www.cablinginstall.com/home...-patch-cords-affect-your-networks-performance

https://www.cablinginstall.com/conn...tchcord-conundrum-corrosion-versus-conformity <--- This is my favorite article on the subject and the main reason I don't use patch cables with injection molding boots.

https://www.cablinginstall.com/cabl...ing-shows-offshore-patch-cords-fail-miserably

https://www.cablinginstall.com/cabl...sessment-of-offshore-cat-6-copper-patch-cords

https://www.cablinginstall.com/cable/article/16469313/patch-cords-prove-to-be-networks-weakest-links
 
As far as what I install it's all Cat6 these days. Price is almost the same as Cat5e.
Makes sense and that's my approach too. But how do we know if we've got a CAT6 system if we can't test it in some way? If we're only testing to make sure that Gigabit Ethernet works then what's the point?
 
Most of my customers get a pin pair test and that is it.

I charge extra to do any type of certification.

I only use name brand cables and have never had a problem.

Sent from my SM-G870W using Tapatalk
 
When I worked for DEC the tools used to troubleshoot/test/certify cable and fiber networks cost tens of thousands of dollars and that was 25 years ago. Only the largest corporate clients demanded and paid for that level of certification.

While on my own like everyone else here my clients got a simple verification test for basic continuity.

https://www.flukenetworks.com/knowl...ation-qualification-and-certification-cableiq
 
So the bottom line is there is no way to know if your CAT6 infrastructure is really installed to CAT6 spec without using a $10,000 test device... and since no one here has done that we just make sure it runs well 1 Gbps and call it a day?

What about getting a couple of 10 Gbps cards and test a PC link point to point?
 
Again unless I am getting paid they get basic testing.

Certification adds time which adds cost which means getting paid more.

I get 1-2 clients a year that require true certification with reports. These are usually large clients such as insurance firms

Sent from my SM-G870W using Tapatalk
 
So a new client has existing CAT6 wires in place that were run previous building owner. What do you do?
 
So the bottom line is there is no way to know if your CAT6 infrastructure is really installed to CAT6 spec without using a $10,000 test device... and since no one here has done that we just make sure it runs well 1 Gbps and call it a day?

What about getting a couple of 10 Gbps cards and test a PC link point to point?

This is an example of theory vs reality. If you look at the actual specifications there is't a single mention, at least that I could find, of an actual test of content transmission. As in transmitting, for example, 1000gb of data and measuring the time to arrive at a transmission rate. All of the test specs are electrical/electronic transmission characteristics.

For example. One has a need to reliably remove nuts from bolts. If the spec developed has a socket wrench with ratchet then an open or box end wrench will not meet the spec. But will work just fine.

I suppose you could use 2 desktops with 10gb nics and do a transmission test. But it'll never come close to the spec because the rest of the system, motherboard, cables, processor, RAM, drive, etc can't handle anything close to that. I think the highest transmission rate I've seen on a computer was around 750mb even though everything was 1gb. Remember that these specs only apply from one end of the cable to the other and not to devices plugged into them.
 
So a new client has existing CAT6 wires in place that were run previous building owner. What do you do?
Unless I'm being paid to do something I don't do anything at all.

If they want to pay to have it checked out we verify everything is cat6 and run a cable check of each drop.

Sent from my SM-G870W using Tapatalk
 
we verify everything is cat6 and run a cable check of each drop
So you visually inspect the patch panel and the jacks to see if they're labeled as being CAT6 then you visually inspect the horizontal runs and make sure that wire is CAT6?
 
That's correct we visually inspect and if they are paying for it test the pin out of each run.

Sent from my SM-G870W using Tapatalk
 
So in a hypothetical reality where the jacks aren't marked CAT5e, CAT6 or whatever, no markings on the patch panel. You can't see the horizontal runs. You're new to the site. How can you tell if it's CAT5e or CAT6?
 
Back
Top