Choosing the Right Termination - Keystone Jack vs RJ45 Connector vs Field Termination Plug
The folks at trueCABLE frequently get asked a great number of questions about the correct cable needed for any particular installation. Up until now, we have gone to great lengths to explain everything about cable jacket types, Category, and whether you need shielded Ethernet cable or not. What also comes up are questions about terminations. This is where things can get a bit more complicated, and how you terminate your cable can vary by installation. There is a defined best way of going about this, and there is also the “it will work” way to terminate your cable. That all said, how do you get the most out of your Ethernet cable and end up with the best “bang for the buck”? How do you know that you will get the best performance possible? Are all terminations equal? That is what this blog will answer. Grab some coffee, as this discussion will be a bit involved and will likely introduce some terms you have not heard before.
Structured Cabling Systems
We cannot really go into what terminations are best without talking about what a traditional structured cabling system is. If you were to walk into a commercial office building and see a keystone jack (female outlet) in every office and then a centralized “server room” where all of the equipment connects up, you just saw a structured cabling system. Structured cabling systems contain:
- Backbone cable running between floors or primary pieces of equipment (often fiber optic these days because it can go longer distances and is very high capacity)
- Copper Ethernet cable drops (also called runs) that are typically used on the same floor that run through the walls and above drop ceilings
- Outlet connectors mounted in or on walls (called keystone jacks) that are used to patch-in (connect) to the network with a patch cable
- Centralized patch panels where all the cable on the floor runs back to, so it can be patched into central Ethernet switches
- All of the support structure necessary to hold the cable in place like ladder rack, cable tray, conduit, and more
- A very defined set of rules and guidelines about how to run the cable like maximum length, amount you can bend it, how hard you can pull it, and more
- Extensively defined ways of labeling and documenting your installation
- …and much more
Of course there are as many variations on this theme as there are installations out there. No two are the same. Even homes that have Internet access with Ethernet cable of any length contain a small-scale structured cabling system that is not all that different (conceptually) from a commercial one. This even includes a single device setup like an “all-in-one” cable modem, router, and WiFi hotspot. It might be small and compact, but the idea remains the same. You have a centralized point for your network and it radiates outward with WiFi and maybe even Ethernet cables. There's probably patch cables you bought from a store, actually. You know, the one you hooked from your combo Internet “thingy” into your TV?
So, considering there is a generally accepted way of running and organizing Ethernet cable there is also a generally accepted way of terminating Ethernet cable correctly, and it can vary by what you are doing.
Terminating Ethernet Cable
If you have the notion that Ethernet cable has something on both ends you would be correct! You have to terminate the cable at both ends or it won’t work. Terminate simply means to make a physical/electrical connection with a connector device of some kind, so you can plug something into it or plug it into something else. Likely, you have a picture in your head that says a termination looks like this:
The common 8P8C Plug (aka RJ45)
In the picture above, you see the most visible type of connector known generally as a RJ45 connector. Actually, it is really a 8P8C (eight position - eight contact) connector but people generally refer to it incorrectly as “RJ45”. We will stick to the incorrect way for simplicity. If you are wondering why this piece of plastic can cause such naming confusion see What is an RJ45 Connector?
If you ever ran to the store or shopped on-line for a 10 foot long “Ethernet” cable that had both of these plastic connectors at both ends then you bought a patch cable. The idea is you can use this patch cable to plug in your TV to your router/switch (in other words, you patched in). This is probably what your idea of how all Ethernet cable should look like, and if you came to trueCABLE’s website and purchased a big roll of unterminated Cat6A Unshielded Riser cable you also likely assumed you have to use the very same connector at both ends. Well, that is not only wrong it is actually the least desirable way of doing things. More on that later.
You may have also run into something like this, and were not entirely sure of what its function was:
A typical keystone jack (usually seen mounted in a wall face plate)
If you look closely, you will see there is a port on the front of the jack. That is what you will see when the Keystone Jack is mounted in a wall. As it happens, that is where a RJ45 male plug goes….
Another doodad that is used in structured cabling systems (but perhaps less commonly) is called the Field Termination Plug, and that looks like a cross between a RJ45 and keystone jack of sorts:
Shielded Field Termination Plug
Many people do not know what field termination plugs are, and what they are used for. It terminates like a toolless keystone jack, but ends up being an RJ45! Weird, huh? Well, you will find out pretty quickly just how effective field termination plugs are and why they are so desirable in short order. Keep reading!
Finally, there is the patch panel. Patch panels won’t be discussed at length here, but suffice to say they are keystone jacks on steroids. In effect, a patch panel is anywhere from 8 to 48 keystone jacks integrated into one big rack or wall mountable unit. They can come in punch down or tool-less variety. This is the centralized attachment point for all of your cable runs and is often mounted in a rack, wall enclosure, or otherwise in a spot out of sight. Here are two examples:
Small eight port punch down patch panel, made by Legrand for their brand of enclosure
12 port tool-less patch panel, where you can snap in any brand standard keystone jacks
Which Kind of Termination Should You Use?
In the previous section, I presented you with a number of examples of ways to terminate the ends of your bulk Ethernet cable. Which one is best? Well, the discussion about “best” comes down to what you are doing, how professional you want to be, and how well you want your cable to work.
If talking formal Structured Cabling Systems, then the choice is quite clear, you should use the “rack to jack” approach for your installation. What the heck is that? This means patch panel to keystone jack or keystone jack to keystone jack. You then use patch cables at both ends of the permanent link to patch in your stuff. Once you plug in the patch cables, you create the overall channel. We need to define some more terms again, and this time with pictures and diagrams to help illustrate the idea.
Very simple rack to jack diagram showing Permanent Links (blue) and overall Channels (blue + grey at both ends)
What does this look like modeled in miniature?
Two patch cables, two keystone jacks, and a solid copper cable in the middle. This is known as a Channel. Channels always end in male plugs. ALWAYS.
If we are talking about two keystones with solid copper cable between them but without patch cables then this is known as a Permanent Link. Permanent Links never end in a male plug. NEVER.
In a professional installation (commercial or pro-sumer residential) this is the type of installation that should be used. In the commercial world, this looks like a big telecommunications room (or rooms) with seven foot tall racks and big patch panels. In the residential world, this likely looks like a 36” tall structured media enclosure with a 12 port patch panel inside. Same idea, just very different sizes involved.
The main thing to underscore here is the patch cables are factory made and purchased based upon needed lengths with already terminated ends. Industry recommendations and best practices dictate you don’t terminate RJ45 plugs except in limited circumstances. Those circumstances revolve around PoE (Power over Ethernet) runs to power up WiFi access points and IP surveillance cameras. You will use a RJ45 male plug at one end only. The other end is either a patch panel or keystone jack, patched with a premade patch cable into a PoE switch.
For more on how factory made patch cables are different from solid copper structure Ethernet cable and what to use where, see What is an Ethernet Patch Cable?
This type of run is known as a MPTL (Modular Plug Terminated Link) and there are two ways to construct one:
- RJ45 8P8C plug
- Field Termination Plug
What does this look like? Well, we have another miniature mockup to show. Actually, two of them.
MPTL run with Field Termination Plug
MPTL run with RJ45 8P8C Plug
The commonality between these two MPTL runs is both use keystone jacks at one end and some type of male plug at the other. The idea is to plug one end into a device that needs data and power over the same cable, hence PoE.
Here we have the two MPTL runs constructed into Channels. They are now Channels because there is a patch cable on one end and the other end is a male plug. Remember, Channels always end in male plugs. ALWAYS.
The field termination plug advantage. Although both of the above MPTL runs end up operating the same way, the field termination plug is much easier to work with. The reason is the field termination plug is not fitment sensitive and accommodates a much larger variety of Ethernet cable thicknesses. Plain old RJ45 8P8C plugs are extremely fitment sensitive, and this is discussed more below. Of course, the field termination plug might be too big for your application and not fit inside of a camera or WiFi access point housing, which would force you into using a RJ45 8P8C plug. It happens.
The “Other” Way…
The least desirable way to construct an overall Channel (that being male plugs on both ends) is to use RJ45 8P8C plugs on both sides and make that the entire channel (known as patch as channel). The problems you can create for yourself are multifold:
- Less flexibility. Too much cable? Too bad. Now you have to coil it up somewhere. Too little cable? That will be a headache!
- Less performance. RJ45 plugs have no “secret sauce” inside them to help your Ethernet cable. They are just silly little pieces of plastic with eight golden contacts inside, afterall.
- Fitment becomes a real problem. In order for RJ45 8P8C plugs to work well (or at all) the plug in question has to be a near perfect fit for your cable. How do you know if you are getting a good match? trueCABLE goes through the trouble of testing everything together for compatibility however manufacturing tolerances can still cause problems on occasion. Using a cable from one manufacturer and a plug from another? Good luck with that, unless both manufacturers have documented their products thoroughly. For more on proper RJ45 plug fitment see Selecting the Correct Connector.
Here is a mini mockup line-up of a solid copper Ethernet cable constructed into Channels of various types. The top one is the least desirable unless you bought a premade patch cord.
They are all Channels, in the end, because they all end up with male plugs
That Old Black Magic
What most people do not realize is keystone jacks, patch panels, and field termination plugs all use PCBs (printed circuit boards) inside that are component rated and help impedance match your cable run. These accessories are Category rated too, which means for Cat6 cable you would use a Cat6 keystone jack. That is their “secret sauce”, and the performance data shows it.
RJ45 plugs at both ends, even if nearly a perfect fit, will give less than stellar performance (but likely work well enough since Ethernet technology is forgiving).
In the course of testing cable, I use a Fluke DSX-8000 CableAnalyzer and will demonstrate the performance differences when Channels are constructed in various ways. There is a defined and accepted way of proper Certification in the field and how you test the cable run is dependent upon:
- How the cable under test was terminated (are we dealing with a patch cord, overall Channel, MPTL run, or Permanent Link?)
- The test run may be tight or more permissive depending on the way the cable run was constructed
- The cable Category will define the test as well
Depending on the test run and adapters used, a number of parameters are tested such as cross talk, return loss, insertion loss, PoE capability, and more.
Fluke test results:
The first column shows how I constructed the cable run. The second column shows whether it passed or failed. The third column shows the test limit (tightness) that was used depending upon how it was constructed, and the fifth column shows how well the cable passed. In other words, it shows the headroom available. The higher the headroom the better the results. The fifth column is extremely useful because while many tests will pass some of them pass…better….than others.
Test setup and methods:
- The same cable segment was used for all tests to minimize variability in cable quality
- Cat6 Unshielded Riser cable was used as the test subject
- The Fluke DSX-8000 is in calibration
- The ambient temperature was 68 degrees F which is ideal for full length testing, depending on the test. Permanent Links can be 295 feet long. Full Channels can be 328 feet when patch cable(s) are added. Yes, temperatures can affect your permitted lengths drastically. See Temperature's Effect on Ethernet Cable Length for more information.
Key takeaways from the above data:
- As predicted, the best performance was seen with a Permanent Link (jack to jack which is equivalent to patch panel to jack for our purposes). This was the test that came back with a 7.8 dB NEXT headroom number and the Permanent Link test is a very tight test.
- Next best performance was obtained with the MPTL runs, which involved at least one keystone jack in the mix. The test used (Cat6 MPTL) is just as tight as a Permanent Link test. The headroom fell a bit since a RJ45 of some type was used at one end of the cable run.
- Worst performance was seen when the full cable run was terminated at both ends with plain old RJ45 plugs. It failed on Return Loss, which is not unexpected. A minor fitment variable caused it, on a cable and plug that have been tested before and passed. It is a hit or miss proposition. The test I ran was the tightest of all…a PCA or Patch Cord test.
- That said, the same “patch as Channel” RJ45 to RJ45 terminated run that failed was tested again but this time on a far more permissive Channel test. This test assumes the cable run will be plugged from one powered device directly to another (like switch to TV, for example). In this scenario, it is OK to do this even though the stricter test failed. However, that is the ONLY scenario where it will work OK. Introduction of any other component might cause a failure.
So, as you can see there are a number of ways to construct an Ethernet cable run and depending on how you do it you may end up with wildly different results. Knowing the most desirable way to terminate cable for best performance and reliability (not to mention flexibility) will pay off in the long run!
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.