Cable Testing

There are two main drivers for your clients requiring their cable installations to be tested.

Cable testing can be driven by the cable manufacturer because they need to ensure that any cabling that carries their guarantee is installed correctly. The manufacturer can then know that any subsequent failures are genuinely as a consequence of in-service failure rather than being due to incorrect installation.

Cable replacement is both expensive and difficult post installation. However, the potential cost of failure due to a fault may be much more than the cost of installation. The client therefore must ensure that the cabling is installed correctly and works as expected.

The different drivers for cable testing also shape the market for cable testers. Cable testers used for certifying cable to manufacturer’s standards tend to be the most expensive, and require constant re-calibration.

Cable Tests

Cable tests can be very simple like the open pair test or extraordinarily complex like the Power sum tests (see below). Each test attempts to measure the performance of your cabling so that you can be sure that it will perform adequately for your customer.

Wire Mapping

Wire mapping is the simplest testing you can do on a twisted pair cable. It is also the most fundamental. If the cable isn’t wired together properly it will struggle to perform to acceptable levels, or perform at all in some cases.

• Open Pair – occurs when one or more conductors are not connected to a pin at either end of the cable. The conductivity of the cable is impaired and will utterly fail to perform in service.
• Shorted Pair – occurs when one of the wire pairs shorts to each other at some point along the length of the cable.
• Short between pairs – occurs when one of the wires from a pair shorts to a wire on another pair at some point along the length of the cable.
• Reversed pairs – One or more wires are not connected to the correct pins at the other end of the cable. A Reverse Pair is when the correct pins are used but the connectors are reversed.
• Crossed pairs – Crossed pairs are when a pair of wires of one colour are connected to pins of a different colour at the other end, so you have to make the same mistake at both ends of the cable. As a result this is a pretty rare fault.
• Split pairs – Pairs should be connected (T568A or B standard) in the order 1&2, 3&6, 4&5, 7&8. A cable would pass a connectivity test if both ends were incorrectly connected with the wrong pairings. If the same mistake is made at both ends than the cable will likely pass connectivity tests.

Cable Length

Specifications exist for all types of cabling and vary depending on the cable type you use and the type of network you plan to install. The maximum allowed cable lengths are an important consideration in your initial planning of the network and you should never exceed them. A network tester should allow length testing.

Although on paper your cable lengths may be fine it is worth checking them on the finished installation. Remember that due to the twist in the cable the actual length of the conductors will be more than the outside length of the cable. Not only will you know for certain that everything is within the specification, but you will also be able to detect shorts or faulty connections and so on by measuring the cable lengths.

Performance Testing

• Attenuation – All signals degrade when transmitted along cables. This degradation is called Attenuation. Too much attenuation can cause signals to be lost and is especially critical on high speed networks and where cable runs approach the maximum allowed for the network type.
• Near-End Crosstalk (NEXT) – Near End Crosstalk (NEXT) occurs when a signal on one pair is picked up by adjacent pairs in the cable. For completeness measurements are made for each pair against the other three pairs and from both ends of the cable. The worst case measurement indicates the performance of the cable.
• Power-Sum NEXT (PS-NEXT) – PS-NEXT is a measurement of the cumulative effect of crosstalk on each wire pair whilst the remaining pairs are transmitting data at the same time.
• Attenuation to Crosstalk Ratio (ACR) – ACR is not technically a test of its own, it is a measurement of the difference between the attenuation for the cable run and the crosstalk.
• Far-End Crosstalk (FEXT) – Far End Crosstalk (FEXT) is a type of crosstalk that occurs at the other end of the cable. In order to average the crosstalk over the whole cable run you need to calculate the Equal-Level FEXT (ELFEXT). Subtracting the Attenuation value from the FEXT value gives the ELFEXT.
• Propagation Delay and Delay Skew – The time taken for the signal to reach the far end of the cable is the Propagation Delay. In twisted pair cables the length of the conductors are not the same, due to the different twists of each pair and so the propagation delay can also vary. On high speed networks that use all four pairs (such as Gigabit Ethernet) the signals need to arrive within a short time period, otherwise data will be lost. The difference between the highest and lowest Propagation Delay in a cable is known as the Skew. On high speed networks that use all four pairs Skew should be as low as possible. The shortest pair in the cable is used as the reference for the Skew measurement.
• Noise – One element that can effect the performance of the cable is not the cable itself or how it has been terminated but external factors like the amount of electrical interference encountered along the length of the cable.
• BERT – A BERT (Bit Error Rate Test) dispatches millions of live data packets on the cable in both directions (full duplex) at the maximum data throughput possible (98.5%) and reports the error rate. This test guarantees the ability of the cable to perform at a specific data rate. The specification as stated in IEEE 802.3 for 1 gigabit data speed is not more than one error in 10Gbits transferred.

Cable Testers

The types of cable testers can be broken down into three broad categories: certification, qualification and wire map testers. The categories are not hard and fast and some overlap exists.

Certification

Cable certification is largely driven by cable manufacturers insisting that cabling they are required to guarantee is properly installed and tested to their requirements.

If certifying the cabling is in the service level agreement (SLA) then you’ve got to do it or you are putting the contract at risk.

Your clients may include a requirement to certify in order to satisfy the cable manufacturer’s warranty requirements.

Cable certification testers are manufactured by the large cable test players like Fluke, Agilent and Ideal Industries.

Unfortunately, they tend to be expensive, costing upwards of £4,000 each and require periodic recalibration and consume expensive connectors.

The good news is that, whilst they are expensive, they are very fast. For instance, the Agilent Wirescope PRO can perform a full cable sweep in around six seconds. Cheaper testers perform much less rigorous tests much more slowly.

If you are installing thousands of cables and are required to test each cable then a cable certifier may well justify its high price in the time saved performing each test.

The two most common cable certification testers are the Fluke DTX 1800 and Agilent Wirescope PRO. Both are available from OPENXTRA Limited.

If you can’t justify a certification tester you can hire one on a daily or weekly basis from a number of suppliers.

Whilst you are very unlikely to come across this, certification testers are the only testers able to test for alien crosstalk. That is, crosstalk between cables in a bundle, rather than crosstalk within the pairs inside a single cable.

With the emergence of 10 Gb (10 billion bits per second) networks in larger installations, testing for alien crosstalk has become necessary.

Qualification

Whilst certification testers approach cable testing from the electrical properties perspective, qualification testers approach the cable from a networking perspective.

Qualification testers are usually used post installation by the users of the network to verify that the network is operating properly.

Most qualification testers implement tests like the BERT test that generates real network traffic and checks that the traffic is being sent and received properly.

Whilst the correct sending and receiving of network traffic should suggest that the cable is correctly wired, in practice things aren’t quite as simple as that.

Cabling, like CAT5e and CAT6 has a large amount of resilience built into it. So, even when things are incorrectly installed, the network may well still outwardly operate correctly.

The most common qualification testers are the Test-Um Validator (and Validator-NT with network tests built-in) and the Fluke CableIQ.

Qualification testers cost substantially less than certification testers and typically do not require periodic calibration or expensive consumables.

Qualification testers do provide an easy way to label your network, typically supplied with PC software for planning testing schedules.

Wire Map Testers

The most basic cable tester is the wire map tester. The wire map tester’s sole purpose is to validate the proper termination of your cabling as per TIA568A/B.

Wire map testers, as the most basic testers and also the cheapest. Prices start from around £50 for the most basic tester.

Test-Um manufacture a range of affordable wire map testers. We have the full range available from stock.