Cable testers have always been an example of extremes. At one end there are the low cost wire map testers which simply check for short and open circuits, crossed and reversed pairs. Useful for fault-finding but giving no indication at all of the performance of installed cabling. At the other end are the cable analysers/certifers which measure all the important physical characteristics of a cable but cost several thousand pounds. In the middle there are wire map certifiers - essentially wire map checkers with TDR functionality added to indicate which end faults are at and the ability to store test results to generate reports or certificates. Again these are very useful for fault finding but give no indication of cable performance or quality of installation. In order to certify an installation the installer would have to invest a considerable sum in a cable certifier or hire one on an ad hoc basis.

Recently a new breed of testers has appeared on the market which occupy the middle ground in terms of cost. Examples of these new cable qualifiers from Test-Um, Fluke Networks and Ideal Industries use Ethernet Bit Error Rate Testing to offer a more relevant test for the network manager at a fraction of the cost of traditional cable certifiers.

Historically cabling installation standards concentrated on formulating the correct methods for installing cables and establishing levels of performance which led to the performance levels of Cat5, Cat5e, Cat6 etc but basing everything on frequency. Traditional cable certifiers use frequency sweep signals to measure conditions that are likely to cause a failure on a cable and compare these measurements against the predefined limits in the test specification. The Ethernet specifications of IEEE802.3 on the other hand defined speed standards at 10Mbit, 100Mbit and Gigabit and it was Gigabit that made life difficult as it uses all four pairs simultaneously despite having the same frequency spectrum as 100Mbit Ethernet systems. This lead to the extra requirements for measuring Power Sum Near and Far End Crosstalk which add in the influence of the other three pairs instead of just the one pair under test. These measurements still only show a condition that is likely to cause a failure and infer what would happen at a given speed.

What's the Difference?
So how do the new testers differ from traditional cable certifiers? Where traditional testers use frequency sweep signals the new cable qualifiers generate digital signals to measure actual noise levels on installed cable, measure signal to noise ration (SNR) and uses the Bit Error Rate Test (BERT) to show transmissions of actual digital data and error rates. Where traditional cable certifiers measure Far End Crosstalk and Power Sum Crosstalk, the BERT test is a true measurement of real data flow and error levels. After all, if you're going to implement Gigabit Ethernet, shouldn't the network be tested with Gigabit Ethernet?

Testers such as the Validator from Test-Um show Signal to Noise Ratio and Skew, both of which derive from the requirements of TIA568, and give an initial indication of any physical problem likely to affect data flow. The SNR represents a combination of elements from all four pairs which contribute to noise for Gigabit, the higher the SNR the better.

Skew shows any delay in the signal propagation along any particular pair to the other end compared with the other pairs. High Skew levels will result in data loss and retransmissions. Excess noise and signal delays are frequently caused by poor terminations, bad installation practice and poor quality components. But that isn't the end of the story. The new cable qualifiers now go on to use the BERT test to show the real functionality of the cable, using actual digital data to measure actual performance.

A Qualifier not a Certifier
Although the new testers adequately demonstrate the data carrying capability of a cable, the fact that they are often referred to as cable qualifiers or speed certifiers, not cable certifiers indicates the stranglehold that installation standards such as those specified by EIT/TIA or ISO/IEC still have. The belief is that these cabling standards are the only measure of how well an installation has been done. This is reinforced by the fact that cable qualifiers can't be used to validate a manufacturer's warranty.

So who will use such testers? They are ideal for installers working on smaller installations where certification or a manufacturer's warranty is not required. They are also perfect for Network Managers who need to check that their own network is running correclty, especially after any moves, changes or prior to any forthcoming cabling work.

By using real digital data, cable qualifiers show how well the cabling works in actual network conditions, giving installers and network managers increased confidence that installed cabling works correctly.

This article was originally published in Wired IN magazine.


Back to News