RJ45 connector performance

[UnplugMe]

I recently ran a Cat6 cable ~75' between two switches and terminated with RJ45 connectors rated Cat6 off Amazon. The link was unable to connect at 10Gb speeds. I assumed it was the connectors since they are cheap and Chinese. I never terminate cables like this for permanent installation. Typically the Cat6 gets terminated in to a keystone or patch panel and then a premade patch cord is used to plug in to a switch. The cable used is True Cable brand and the connectors are Cable Creation. I have been trying to convince my employer to start using a more reputable and consistent brand or supplier but have difficulties convincing them that it is the equipment and not the installer(me). Just for reference before my current job I worked for an MSP that had a cabling division and I ran over 200,000 ft of network cable over the last 10 years and the 3 main technicians were all Bicsi certified. I would be the helper when the workload was slow on the IT side but over time I was doing the smaller jobs myself.

We brought down new connectors the next day and tried each one out. My employer brought some no name pass thru connector purchased off Amazon and I brought Simply 45 brand Cat6 connectors. We tried both connectors and not surprisingly the pass thru connector didn't do any better and my Simply 45 connectors did work at 10Gb. Now he has decided to keep buying cheap connectors and keystones from Amazon and when we need 10Gb connections only go fiber.

Does anyone have actual cable certifications to show the difference in cable quality and connector quality? I have a hard time convincing them to change without hard data. I have approached a local vendor and they don't have any data besides the datasheets from the cable manufacturer which they all look the same. I am really looking for a cable qualification test of a premade cable vs DIY and tests with different connectors.​

 

[Scott24x7]

My first question is... have you proved that the gear at both ends are actually communicating at 10GB? If they are not, then the cable is not the issue, and you're focused on the wrong thing.
You can also get a cable tester and check it before hand. If you have a known cable tester with a known cable that performs at 10GB, and you connect that same tester to your 75 foot run, and then it doesn't perform, then you know it is either the cable (which you also haven't eliminated) or the connector. So for the moment, your tests have proved nothing other than you have a 1gb connection.


Best Regards,
Scott
MSc ISM, MIET, MASHRAE, CDCP, CDCS, CDCE, CTDC, CTIA, ATS

"I try to be nice, but sometimes my mouth doesn't cooperate.

 

[UnplugMe]

We have not verified transmission at 10GB. I can see both switches auto-negotiate to 10GB but we only have one device that has a 10GB adapter. I can use iPerf tonight after hours to verify the link will pass that much traffic.

I guess my bigger overall issue with this is I have always had questions with the quality of the cable and connections that come off of Amazon and this is just the first time it obviously became a problem. Are these components causing problems I don't know about? The switch doesn't show dropped packets or anything else that I would expect but they have hundreds of cables deployed with these connectors and various other brands of cable including some CCA. Should I be worried that they really are not able to perform at 1Gb reliably either? I only have a continuity tester and don't know anyone that has a cable qualification tester to look at the actual cable performance.

 

[Dexman]

I'll second what Scott24x7 said.

Purchase a premade patch cord (75' or 100') and use it to connect both devices together. If the units can connect @ 10GB, then the problem could be either the cable, the connectors or both.

My personal preference is to purchase premade patch cords, but, I've made patch cords using Systimax 2091 cable & Platinum Tools EZ crimps without running into any serious problems.

I "FEATURE 00

 

[TouchToneTommy]

Sorry, but I gotta ask, what is the color code that you are using when you make these cables?

 

[Scott24x7]

Another thing you could do is get a higher quality cable (like a Cat-8 or a Cat-6A [but not Cat-7]) and connect it, and see if you get the same performance. If you do, then it's the gear and not the cable that's performing at a particular capacity. I have, on rare occasions, seen issues with the quality of a cable. But typically if you're using same-rated connector to cable, there aren't many issues. The quality is usually not with the conductor, but with the casings they put around them.


Best Regards,
Scott
MSc ISM, MIET, MASHRAE, CDCP, CDCS, CDCE, CTDC, CTIA, ATS

"I try to be nice, but sometimes my mouth doesn't cooperate.

 

[allworxguy]

Why have you not run fiber? Much more reliable at that speed.

Marv ccna

 

[budbyrd]

As a personal opinion. I would put jacks on both ends and use factory made patch cords.

Dermis and feline can be divorced by manifold methods.*
*(Disclaimer for all advise given)--'Version Dependent'

 

[libellis]

I'm in agreement with what budbyrd said. Field terminating standard Cat 6 (or possibly sub-standard in this case) RJ45 plugs for a link at 10Gbps is not really considered an acceptable practice. There are plugs suitable for doing this (and Cat 6A recommended), but not the plug you purchased. For the distance of your link, terminating on a jack and using good quality factory patch cords will give a better chance of working (assuming the cable is good quality).

 

[libellis]

Forgot to add - the jacks obviously have to be good quality - otherwise you're back to square one.

 

[Commshark]

Hello, just wanted to chime in. The brand of cable you are using (True Cable) is probably one of the best cable manufacturers you can get on Amazon. It is ETL listed and they Fluke test all of their cable. They also don’t offer any CCA crap.

http://www.commshark.com