STP and 10Gb/s: a marriage made in heaven

First system tests in live conditions

There is now conclusive proof that shielded copper is the ideal cabling infrastructure system to supply 10GBASE-T and larger applications.

This news resolves the old debate over the pros and cons of STP and UTP systems. Until now they have been seen as two alternatives, each with their own supporters, each preferred in particular geographic areas. It seemed to be a matter of tradition and preference, with neither having a clear overall advantage.

But now that has all changed. The world has moved forward, new issues have become increasingly important and STP systems have emerged as the clear winners.

What are the changes that have led to this? Probably the most significant single change is the growth of 10Gb/s and the new demands that this places on our systems. And at the same time we have seen growing concern that systems must be suitable for the type of environment in which they are installed. This has resulted in the definition of new standards that have far-reaching effects on how we judge the suitability of different systems. These are the international MICE classifications.

As well as the big gap in data transmission speeds, there are two major differences between 1Gb/s and 10Gb/s. Firstly, due to its complex modulation encoding, 10Gb/s operates with much lower signal levels per symbol, and these are more likely to suffer interference from background noise. Secondly, since 10Gb/s cabling uses much higher transmitting power it also radiates much more, especially between 250 and 500 MHz; the result is that neighboring cables interfere with each other, an effect called Alien Crosstalk or ANEXT. So it is important to understand which type of system performs well in relation to these issues.

The breakthrough in our knowledge has been the recent GHMT testing. Before these tests, the information tended to emanate from the manufacturers of the systems. So whilst it was becoming more important to have a clear view of the advantages of each system, objective evidence remained hard to find. A neutral expert had to be found who could resolve the issues.

HMT are a vendor-independent organization, so they satisfied the need for impartiality. Established in 1992 in Germany, they have become internationally known as a leading ISO/IEC 17025 certified laboratory, specializing in the testing of cabling and connection components. They were chosen to carry out tests on a variety of shielded and unshielded systems. The value of these tests was enhanced by the fact that they included real 10GBASE-T equipment in environmental conditions as defined by the international MICE classifications.

The significance of MICE is shown by its inclusion in the world’s standards: ISO, CENELEC and TIA. The MICE classifications are Mechanical, Ingress, Chemical and Electromagnetic. For each classification there are three levels, which represent the different degrees of environmental control prevalent in different areas. The pass/fail criterion used by GHMT was Level 1, which applies to offices and similar areas, but they also measured against the higher levels, relating to uncontrolled environments.

Test set up with live 10 GB/s application.

The tests were to be carried out with six Class EA cabling systems from six different market-leading suppliers: 3 x U/UTP, 1 x F/UTP and 2 x S/FTP. The test criteria were to ISO/IEC 11801 Ed. 2.1 Class EA channel.

The first finding was that STP systems offer high coupling attenuation and therefore extraordinary ANEXT performance. In fact one of the UTP systems failed in the PS ANEXT parameter and had an unacceptably low coupling attenuation, so it did not qualify for further MICE testing and had to be withdrawn from the study. Coupling attenuation is the key parameter for measuring the EMC performance of twisted pair cabling systems.

It was clear that the design of STP systems is a good match for the requirements of 10GB/s. Unlike UTP systems, STP systems do not require metal conduits, need no separation distance from power cabling, can mix 1 and 10GB/s in one conduit without ANEXT disturbance and do not need mandatory ANEXT testing.

The focus of the tests was the five electromagnetic MICE parameters, which are radiated high frequency, conducted high frequency, fast transient, magnetic field and electrostatic discharge.

Test result for electromagnetic wavelength.

The first of these, the radiated high frequency test, determined each system’s immunity against electromagnetic fields from 80 MHz to 2 GHz. The test simulated potential sources of disturbance such as mobile radio installations, local radio or broadcasting stations, mobile phones and walkie-talkies. The pass level was 3V/m, and the STP systems passed for all levels, i.e. all electromagnetic environments. The U/UTP cables failed, falling well below even MICE Level 1, the lowest level. Just pushing the push-to-talk switch or the call set-up button was enough to impair or even stop the data transmission of unscreened systems.

Conducted high frequency was the second parameter to be tested, checking system immunity against magnetic fields from 160 KHz to 80 MHz. Again, potential sources of disturbance were simulated and again the results were clear: STP passed for all electromagnetic environments, whereas UTP failed for all.

Practical test with mobile phone.

The next tests measured the effect of fast transient electrical disturbances caused by switching. Again, all round success for STP and failure for UTP systems. This time there were also practical tests, firstly with fluorescent lights in close proximity. Impairment or even complete breakdown of data transmission resulted for unshielded systems. The systems were then tested when installed adjacent to power cables and the STP systems passed all MICE tests even when there was no separation, whereas a 30cm gap was needed for the UTP systems.

The fourth test focused on the magnetic fields from sources such as generators, busbars and low voltage transformers. This was the only test in which not only STP but also UTP systems passed for all electromagnetic environments.

Finally the systems were tested for immunity against electrostatic discharge caused by floor coverings, clothing, etc. For both air discharge and contact discharge, STP systems passed and UTP failed for all electromagnetic environments.

So the debate is now over: standards based testing by a neutral internationally respected laboratory has proved that shielded copper systems are the logical choice where there could be a need for 10GB/s data transmission. In summary:

• UTP systems failed in all the MICE tests for 10Gb/s.
• UTP systems need special installation guidelines in order to meet ANEXT and MICE requirements
• STP systems meet the radiation standard.
• STP systems passed all the MICE tests for 10Gb/s.
• STP systems meet ANEXT by design.

The performance of a twisted pair cabling systems is from now on:

                                           Transmission performance + EMC performance

More information can be found at: http://www.utp-vs-stp.com/web/Microsites/UTP-vs-STP/.