Ethernet Couplers - How Many is Too Many?
Written by Don Schultz, trueCABLE Technical Manager, Fluke Networks Copper/Fiber CCTT, BICSI INST1, INSTC, INSTF Certified
Have you ever found yourself in a situation when you have two short Ethernet patch cables and need to make a single longer one? If this is the case, you likely picked up Ethernet couplers for home use. At the very least, you have probably seen them for sale. If you work in a large corporation, your IT department will likely yell at you for using one. Used judiciously, they are a lifesaver and get the job done. Use too many, and you will ruin your signal! Grab a cup of coffee and come along with me as we talk about an object that is loved by many and similarly denigrated and despised by those who have had bad experiences with them. I even came armed with Fluke DSX-8000 test results as I had questions too!
Hero or villain?
Why is such an innocent looking connector “doo-dad thingy” such a potential source of controversy and trouble? Quite simply, it is because it is easy to use too many of them, and termination quality matters just as much or even more. But first let’s step back and define what we are talking about…
What is an Ethernet Coupler?
A picture says a thousand words
Quite simply, you have two cables that end in 8P8C (RJ45) plugs, and you need to join them. You cannot cut off the ends and splice them together, as that is not allowed per the specifications and will ruin your performance. The only option is to join them together with a passive coupling device (or use a powered Ethernet switch, but that might not be feasible for you).
Ethernet couplers have the following characteristics:
- Have female ports on both sides of the coupler to join two RJ45 plugs
- Are Category specific, such as Cat6 and Cat6A like trueCABLE offers
- May be inline or keystone jack in design. Inline couplers do not snap into a wall plate (also called faceplates). Keystone jack couplers snap into faceplates or keystone patch panels. Keystone jack couplers “kinda sorta” work like keystone jacks in that sense.
- Come in shielded and unshielded depending upon whether your Ethernet cable is shielded or unshielded. It is OK to use unshielded Ethernet cable with shielded couplers. Do NOT use shielded cable with unshielded couplers as you may create an opportunity for interference on your cable run.
Keystone coupler (left) and inline coupler (right)
There are also outdoor waterproof inline couplers
What Do Professional and Industry Standards Say About Ethernet Couplers?
The conventional wisdom of the professionals is to not use them. These devices are not technically recognized per ANSI/TIA 568-2.D. That all said, if the coupler is built properly (and many are not), then you can get away with using them judiciously and sparingly. I have Fluke DSX-8000 test results that prove it, but there are two important caveats. The first is using a quality coupler. Most manufacturers and resellers will provide quickly assembled “channel level” couplers that do not help impedance match your cabling. trueCABLE, on the other hand, is providing component rated and impedance matching couplers. If “component rated” vs “channel rated” is something new to you, then I suggest you take a look at Ethernet Quality: Channel Level vs Component Level Rating.
The second caveat is the quality of your terminations to start with. Your termination quality can make or break your entire cable run, whether you have a coupler in there or not. A “sus” termination or “sus” coupler can combine to ruin your day.
Remember, the very definition of a patch cable is how it is terminated--RJ45 male connectors on both ends. The least best way of terminating Ethernet cable is plain RJ45 8P8C plugs on both ends, especially if you are doing it yourself and onto solid copper Ethernet cable. There are quite a few things that can go wrong. If you wish to construct patch cables yourself, use Field Termination Plugs at both ends. Another option is to buy component rated and high quality Ethernet patch cables (typically using stranded copper conductors) that really do pass the proper Category testing.
As it turns out, the quality of the RJ45 style terminations matters just as much as the quality of the coupler!
How Do Couplers Get Used?
Typically, the use of an inline coupler is “ad-hoc” which means you did not plan for it. It may not be permanent, either. Keystone jack couplers, however, are designed to snap into wall plates and patch panels and may be used to construct a permanent in wall cabling system using patch cables.
In short run scenarios, and if you don’t have the tools or desire to terminate solid copper permanent structure cable, you can construct a professional looking and tidy Ethernet network using pre-terminated stranded copper patch cable. You can also get into trouble with this strategy, so be sure to keep your lengths less than 125 feet (total end to end run length) for best signal quality and PoE compatibility, and use 26 AWG or 24 AWG stranded conductor patch cable.
Wallplate, where a keystone coupler (or jack) would snap in
How Many is Too Many?
Ok, so we have talked up to now about what couplers are and how they are used. You also know that couplers join up cabling that has RJ45 (or RJ45 like, in the case of Field Termination Plugs) connectors. Testing will prove a challenge. In the mix you have cable quality, termination quality, and coupler quality.
You have to eliminate variables!
I conducted extensive coupler testing by using trueCABLE prototype patch cables. Each and every one of these patch cables was required to pass ultra-strict Fluke DSX-8000 Patch Cord testing using PCA adapters (patch cord adapters that are Category specific, because they measure crosstalk at the connector). I had to be sure the patch cords fully passed before I could start the coupler testing, since I had to be sure of what was being tested. This effectively eliminated the cable and termination variables, to a point. As for coupler quality, I relied upon factory Certification test results. The couplers are subjected to Category component level testing before, during, and after construction. The PCB (printed circuit board) found inside our couplers has to independently pass testing too.
- Multiple prototype trueCABLE Cat6 and Cat6A patch cables were used, in both shielded and unshielded versions, depending on the couplers under test
- All patch cords are constructed with 26AWG stranded pure copper conductors
- All patch cords are factory pre-terminated
- All patch cords had to pass PCA Category testing first
- An in-calibration Fluke DSX-8000 was used for all testing, and the tests conducted are the strictest possible for the given circumstances
- Ambient temperature was 68 degrees F
- A great big pile of couplers
- A big pot of coffee
Tests were conducted by inserting a coupler, testing, and then inserting another coupler and testing, and so on. Results were recorded.
Above is the Fluke DSX-8000 test results summary. I kept going until the tests started failing. I apologize for the extensive acronyms in the Cable ID field. There is only so much room I can use.
- KYCPL = keystone coupler
- INLNCPL = inline coupler
- STP = shielded coupler
- UTP = unshielded cable
- FTP = shielded cable
Some generalizations can be made from the above table:
- Three couplers or less pass testing, but there is one exception that achieved a marginal pass with three couplers (technically a fail)
- The moment you reach four couplers, you are done
- Cat6A testing is more sensitive to adding couplers, which is no shock
- The results are a bit misleading at first glance! As it turns out, it is not the number of couplers that will kill your channel. It is the number of RJ45 terminations (or just terminations in general) increasing with each coupler added that actually ends up in a failure. Academic, I know. In order to use that many couplers you have to have that many terminations.
Just out of curiosity, what does the “picture” of the test look like when things blow apart? Well, let’s take a look at three test results. The first you can consider a “baseline” which is a Permanent Link test that goes from keystone jack to keystone jack and has only two total terminations.
This is a measure of crosstalk. This is a Cat6 keystone jack to keystone jack Permanent Link with no couplers. Punch down termination at both ends. See how nice and behaved it is?
Just one coupler inserted into the Cat6 Permanent Link, but still passing. It is already blowing up. Any more couplers in this Permanent Link, and it will FAIL.
Why did it start to fail so fast? Well, the Permanent Link test is a tight one. You are also dealing with four terminations once you add a coupler, and the test is designed to account for two. Permanent Link tests are held tighter to account for unknown (but assumed good) quality patch cable at both ends of the Channel.
This means we have to loosen up the test limits to Channel if we want to proceed with more than one coupler. When conducting Channel tests, two additional patch cables are added along with yet more couplers and patch cables as the testing proceeds.
What does a Cat6A Channel test look like when you have four couplers in it?
Yikes! Not good. This has totally blown up. This is a Cat6A Channel test with four couplers.
Essentially, when you measure the performance of two couplers together you are measuring an overall channel that has up to six terminations. At three couplers, you are now at eight terminations. At four couplers, you are at a staggering 10 terminations. Well, yeah, no wonder stuff is starting to fall apart! Each termination is a serious disruption in the balance of the cable, electrically speaking.
Coupler Rules of the Road Matrix
It should be noted that these rules assume you have terminated your solid copper cable to Field Termination Plugs at the coupler or you are using good quality patch cables and couplers that pass the proper testing.
Tips for Success:
- Keep your coupler use minimal per the table above, avoiding the use of more than one. Increase to two per the table above, only if necessary.
- NEVER exceed three couplers in a Cat5e or Cat6 2.5Gb/s or 5Gb/s cable channel segment (powered device to powered device) if you want your segment to work reliably
- NEVER exceed two couplers in a Cat6A 10Gb/s cable channel segment if you want to reach 10G at all
- If the quality of your coupler or any component is unknown, restrict your use to a single coupler and no more, regardless of speed
So, there it is. Quite frankly I wondered just how many couplers you could get away with too. If I was wondering about it, I am certain others have too. The answer is not as clear cut as I would like (nothing ever is…) but here is more information to help you out. The basic rule is simply avoid couplers unless you need one. If you need one, use as few as possible (like, one or two at most in the overall run) and be certain of the quality of your components. With that, I will say:
trueCABLE presents the information on our website, including the “Cable Academy” blog and live chat support, as a service to our customers and other visitors to our website subject to our website terms and conditions. While the information on this website is about data networking and electrical issues, it is not professional advice and any reliance on such material is at your own risk.