| ||||||||||||||||
|
EMC
EMC performance is a critical parameter that has previously been understated EMC performance of cabling is a major source of confusion for everyone involved, i.e. both component manufacturers, suppliers, installers and LAN owners. The panic when the European EMC Directive became mandatory has now died down. Not because the major problem of understanding its consequences for cabling has in any way been solved, but only because generally no problems arose when the EMC Directive became effective. This could lead people to believe that EMC and cabling is not critical, a problem of the past and well taken care of by "others". This could not be more wrong. The EMC issue is getting more and more significant for safe transmission as the bit rate continues to rise. Today also application committees have to start to identify the min. EMC performance for high speed cabling like they do for all other traditional and new transmission parameters. To understand the cabling EMC performance issue one has to know the background of the problem and especially realize the fundamental difference in significance of electromagnetic emission and susceptibility of cabling. These issues are discussed below. Background of the problem Cabling and cabling components are passive and consequently no reference to EMC legislation, EMC Directive or traditional EMC Testing can therefore be made. This is easy to understand. To perform an EMC testing on an equipment one powers the instrument(s) and measures both how much it radiates and if it can function safely when exposed to a "normal" level of electromagnetic disturbances. Basically this cannot be done on cables and connecting hardware, and consequently traditional EMC testing of cabling is impossible, and for instance directly excluded from the requirements of the EMC Directive. As always mother nature is a bitch, in this case because the EMC performance of cabling does become significant and part of EMC legislation the second the installation starts running. Now electromagnetic disturbances are generated by the transmission on the channels and likewise electromagnetic disturbances from external sources (PCs, mobile phones, other cabling, etc.) may be absorbed by the cabling. The consequences of the above conflict have been: 1. Cable and Connecting hardware producers could not supply general (generic) EMC warranties or even EMC performance data for their components. Some producers have during time circulated EMC certificates from EMC testing using specific PCs and claiming compliance with for instance CISPR 55022, but such testing has limited value as it only covers testing of a specific equipment in the applied test set-up. These EMC certificates have therefore been a source of much confusion and many disputes. 2. Authorities (EU, national EMC legislation) have denied inclusion of cabling in test and performance regulations. Installations are of course included in the way that PCs, active equipment and cabling are considered one full "unit". 3. System suppliers can perform EMC testing on their total system, but such testing has no or only very little value when new active equipment or PC's are installed by the end user. 4. The application committees had no way to specify the minimum EMC performance of the cabling intended to run the application in question. They might specify the balance, but this would only solve the problem for unscreened cabling and even only if the total effect of balance of cables and connecting hardware was evaluated. 5. The PC and active equipment producers must certify their equipment for EMC performance when connected to relevant cabling. But they are fairly free to define what "relevant" means, giving them a possibility to generate too optimistic EMC test certificates. 6. The poor end user gets all the problems. He knows (or should know) that EMC performance of his cabling could be crucial for safe running of future applications and that he is legally responsible for complying with national and European EMC regulations. For new installations he also has to decide between unscreened and screened cabling based on estimated and incorrectly claimed EMC benefits. Still he has so far not been able to get any lasting assurance of EMC performance for his cabling. The best way out of this dilemma is to know the fundamental EMC behaviour of cabling with respect to emission (disturbances from cabling) and susceptibility (disturbances against cabling). Emission Emission from cabling primarily depends on the PCs applied. Emission levels from active equipment are often close to the limit, and could well be so close that test conditions with respect to mouse and keybord need to be considered. In reality modern Cat. 5 UTP cables give only little further degradation of emission of the system. Hence: It is difficult to argue strongly against UTP with reference to emission. UTP cabling will have lower EMC performance than correctly terminated screened cabling, but this is not very important as the PCs and active equipment most likely will be worse. It should be noted that channel length (cable length) does not have a direct impact on emission. This is contrary to what is later concluded for susceptibility. Susceptibility Susceptibility and emission are directly related in the way that they are reciprocals. This would immediately lead to the thought that when emission is uncritical (see above) then susceptibility could also be ignored. Unfortunately this is completely incorrect. No comparison can be made. When cabling is exposed to a normal electromagnetic disturbance a voltage is generated over each pair depending on the EMC performance of the pair in question. This will depend on the frequency but could typically be from 1-50 mV for Cat. 5 UTP cabling and from 0-0,5 mV for well screened Cat. 5 cabling. In both cases values are generally independent of cable length. When a signal is transmitted from one PC to another the signal is attenuated through the cabling. For short channels with low frequency transmission the received signal will be of Volt level and disturbances of even 50 mV would then hardly present any problem for the receiver. For long channels with high frequency transmission the received signal could well be of 50 mV level, and then a superimposed 1-50 mV disturbance would probably form a potential problem causing slow or interrupted communication. The overall conclusion is that EMC Performance will become increasingly important with higher transmission rates due to the associated higher attenuation of the cabling and/or complexity of encoding. One can be sure that the US based application committees will try hard to allow the future protocols to run on cabling with EMC performance corresponding to well balanced Cat. 5 cabling. However, when performance of cabling is squeezed to the limit it is essential to know and specify the EMC performance to avoid bottlenecks in transmission speed due to electromagnetic disturbances. Coupling Attenuation: A New Method to Identify EMC Performance Awareness about the EMC problems of cabling has forced standardisation committees to develop a method of identifying EMC Performance of cabling and cabling components in spite of their passive nature. The answer is Coupling Attenuation, which is the measurement of EMC Performance in dB. Coupling Attenuation is a measurement of the electromagnetic noise emitted from cables and connecting hardware. It behaves equal to the near end crosstalk, except that here it is not between two pairs, but between a pair and the surroundings. The measurement is easily performed by a surface current sensor provided that the experience in testing is present. Standardisation of Coupling Attenuation is in progress, but significant experience is already now gained concerning coupling attenuation (EMC Performance) values of different constructions of cables and connecting hardware with values as specified below. It should be noted that the below referenced EMC performance ratings are specified by 3P and presently not subject for standardisation. Coupling Attenuation of permanent links and channels is determined by the Coupling Attenuation of the lowest performing component. EMC Performance values for most cables and connecting hardware, and some permanent links and channels are specified in the respective tables in the 3P Survey of Qualified Components. Characteristic Ranges of EMC Performance (Coupling Attenuation).
RATING EMC PERFORMANCE CHARACTERISTIC CABLING |
||||||||||||||||
|
||||||||||||||||