CAT6A cables - Not all patch cords are created equal

The humble yet indispensable Cat 6a patch cord

A quality structured cabling network is the foundation of every high performing IT network. Without the proper supporting infrastructure, devices that transmit and receive large amounts of data at extremely high speeds will be limited in their capabilities.10 Gigabit per second is the highest data rate that can be transmitted over copper cables in structured cabling over a distance of 100 metres. At this speed, it takes about half a second to transmit the content of an entire CD ROM. High-performance switches and servers in blade centres will be the first ones using this leading technology. Therefore, it’s no surprise that high-quality Cat 6a cables and connecting hardware are required for high-speed networks.

However, it may come as a surprise, that in such highly sophisticated networks, patch cords are responsible for more problems in the network than one would imagine and it is often difficult and time-consuming to determine that the problems and network downtime are caused by a simple patch cord. Using cheap, poor quality or non-compliant patch cords in the channel link degrades the performance of the whole channel. A CAT6a patch cord, which is used for connections between switches and patch panels, in between outlets and PCs or other devices, is truly an indispensable part of any network.

 

Unlike 10 or 100 megabits per second networks, high-performance networks like Gigabit Ethernet and 10 Gigabit Ethernet won’t tolerate low-quality patch cords. There is no margin for error and high-quality interconnecting patch cords need to be used. Due to stringent performance requirements for CAT6A components, extra care, testing and quality assurance have become even more important. Component performance classifications defined by the standards may not be sufficient, especially for copper patch leads. At times, you may come across CAT6A patch cables which state on data sheets that they are standards compliant. However, one has to bear in mind that standards allow for tolerances.

A link consisting of cheap components, which just meet the standard’s requirements, will not reach its full performance when the components are not matched and when they do not have sufficient headroom. How will the link perform in a real-world environment? Precise, reliable testing to verify these components are a must, but testing individual components, especially patch cords, is not easy at all.

 

De-embedded and re-embedded testing

Connections are the most critical parts of any IT channel, and the higher the data rate and the frequency, the more complicated it gets. For example, with the old 100 Mbps Ethernet, it was sufficient to roughly set the crosstalk limits of the two twisted pairs. Specifications for the connectors were defined, and the jack had to compensate. With Gigabit Ethernet, this was not sufficient anymore. Specifications for near end and far end crosstalk (NEXT/FEXT) and return loss (RL) for all four pairs of the combination of plug and jack had to be introduced. In order to test accordingly, the de-embedded testing method for Category 5e and Category 6 components was developed between 1998 and 2001. Now, reliable testing was possible in the frequency range from 1to 250 MHz and for data rates up to 1 Gbps. Being backwards compatible with existing Category 5 components was a very important point to cover a large spectrum of existing connectors. As the connectors did not differ too much, a specification range was defined, represented by a set of twelve test plugs.

That sounds easy, but it makes testing extensive and complex.

With de-embedded testing the following steps take place:

• A reference plug is tested using a simplified testing method similar to the old category 5, accepting a relatively large test error. The plug is connected to the test equipment with twisted pairs, using a spe­cial, in most cases pyramidal test adapter.
• The plug and jack are connected to this adapter. The test results of the plug tested before are then subtracted from the results of the plug and the jack (de-embedded).
• What is left are the results for the jack alone, and this jack becomes the reference jack. With this refer­ence jack, the set of twelve test plugs is created to be standards compliant.
• The future test plugs are tested against the reference jack, the results of the jacks are subtracted from the combination of jack and plug (de-embedded) and the results are then compared to the specifica­tions in the standard.
• Only after this extensive procedure of having the twelve test plugs completed, can the actual testing begin: testing a jack against all of the twelve test plugs.

Only when all of the test results with all of the test plugs are within the specifications defined by the standard, the device under test (the new jack) is standard compliant. Only a tested and approved jack may be used to evaluate other components such as patch cords. De-embedded testing can be used for Category 6A (CAT6A) components according to EN 50173 (and the global standard ISO/IEC 11801) with a frequency range from 1 to 500 MHz and data rates up to 10Gbps, but it results in significant time-consuming effort when precise and reliable test results are required.

Because of this, the re-embedded test method was developed, which can be used for frequencies even beyond 500 MHz. As opposed to the de-embedded method, re-embedded testing is very straight-forward. First, a reference plug is tested in a special adapter called “direct probe fixture”. Magnitude and phase of the plug are tested directly with minimum tolerance.

For example, when the NEXT results of the reference plug are known, the combination of the reference plug and the jack can be tested directly. This is a major simplification, which led to the name of this new testing method. As the results of the reference plug are known with minimum tolerance, they can be subtracted from the combination of plug and jack as before (de-embedding).

Then, the results of abstract test plugs specified in the standard can simply be added (re-embedding). Thus one gets the results for all the cases with just one test and the subsequent calculations (in reality, it’s a bit more complicated, of course). The new test procedure is much easier and much more straight forward because the new test adapter was developed without a pyramidal fixture and without additional cabling. It leads to more precise testing of the reference plug and of the combination of plug and jack. The specs for the test plugs (re-embedded testing uses fourteen instead of twelve test plugs) are more precise as well.

A possible reference plug is the commercially available Salsa plug. This printed circuit board based plug can be connected directly to the re-embedded test adapter without the need for additional wiring. However, IEC 60512-27-100 and ISO/IEC 11801 do not dictate the use of this very connector. Telegärtner’s precise, printed circuit board based CAT6A MFP8 connector has the same electrical characteristics and can, therefore, be used as a reference plug as well. This offers an economical solution at the same high quality.

 

CAT6A patch cable standard-compliant test configuration and minimum requirements

The international standard ISO/IEC 11801 Ed. 2.0 Adm 2 was published in April 2010 and specified CAT6A components, but it wasn’t until July 2010 that the corresponding standard for testing – ISO/IEC 61935-2 Ed. 3 – was extended from 250 MHz (Category 6) to 500 MHz (Category 6A). Physics made reliable, precise testing of CAT6A patch cords very difficult.

The standard defines specifications, test configuration and test adapters for CAT6A patch cords. It demands a network analyser as the high-precision test device, and it specifies the test adapter and the test terminator. This test configuration is a so-called two-port configuration, i.e. a configuration with one port in and one port out.

With a two-port configuration, NEXT can only be tested between two pairs. To test all four pairs simultaneously, an eight-port network analyser like the one in the Telegärtner lab (see picture) has to be used, transforming a simple two-port configuration into an advanced eight port one.

 

Testing with or without baluns?

Being very precise test devices used for frequencies up to several GHz, network analysers must have coaxial connectors. The patch cord under test and the RJ45 jacks of the test adapter and the test terminator, into which the cord is plugged, use twisted pairs. To connect the different cable types correctly, baluns are used in the standard test configuration. They connect twisted pair cables (balanced cable) with coaxial cables (unbalanced cable) and adapt them mechanically and electrically, as the two cable types do have different electrical parameters.

 

But like all electrical components, baluns are frequency dependent and so have an influence on the test results. With a self-designed test adapter that exceeds the specifications of ISO/IEC 61935-2Ed. 3, Telegärtner was the first company worldwide who could test CAT6A patch cords reliably and repeatedly. The test engineers could benefit from decades of experience in both, high-frequency technology and coaxial components. Using multi-port network analysers, the Telegärtner lab has been able to abandon baluns for 15years, which is allowed by the standards.

The CAT6A jack used in the Telegärtner test adapter is fully compensated and has more than enough headroom over the specifications of IEC 61935-2.

 

Real-time Re-Embedding

The test procedures of the Telegärtner lab go far beyond the standards and use the eight-port network analyser even for qualifying the jacks. This is the only way to test the interaction of all four pairs of a jack at the same time. The re-embedding calculations are done directly and in real time. With this Real-Time, Re-Embedding Telegärtner has set the new international test standard.

 

3D field simulation for optimal solutions 

The better the transitions between components and wiring are, the lower loss and transmission problems at higher frequencies occur. With test adapters, this means that the better the transitions, the more precise the test results are. Because of this, the electromagnetic fields and waves on all the transitions were evaluated and calculated using the finite element method.

The results and knowledge gained went into the design of the test adapters in an early design phase, leading to adapters that are more precise than the standards demand. Telegärtner developed the adapters and spare jacks with leading test labs along with the standard. They are manufactured in a small series and are commercially available to other companies as test references.

 

Summary

These outstanding results have only been possible because research, development, design and manufacturing have been linked together very closely. These synergy effects have already led to such innovative product details like the patented, integrated protection against over bending the contacts of the RJ45 jack. Eight position RJ45 plugs and six position RJ11 or RJ12 plugs can be plugged into the same jack by turns without damaging the outer contacts of the jack. The consistently high-quality level is documented by the PVP certification of the highly regarded, independent test lab GHMT.

Therefore, in order to ensure high-quality CAT6A patch cords are produced, world-leading testing and manufacturing standards must be in place. Compliance to ISO/IEC, TIA/EIA and EU standards is an absolute must, however, patch cords which exceed minimum standards and provide additional headroom are likely to perform better in real-world conditions. The risk of network downtime, due to inferior quality patch cords is one that can be eliminated by choosing internationally recognised CAT6A patch cords. As a proud supplier of Telegartner CAT6A patch cords, Warren and Brown Technologies offer a complete range of CAT6A patch cords in both shielded and unshielded variations which all come with a 25-year warranty. This offers users peace of mind that the patch cords will outlast the life of the network. In addition, shielded CAT6A patch cords are also available with a moulded angled boot to reduce stress on the patch cord where tight bend radiuses may compromise network integrity.

Drive your network further with Warren and Brown CAT6A patch cords.

Thank you to Telegartner Germany for providing much of the technical content.

12/18/2014