Notes on certification

Independent of the certification device used, the following practically orientated approach is suggested for the certification of a field-configured FastEthernet cabling section for EtherCAT (including end connector test):

Limit values: the use of the channel limit values EN50173-1 Class D for signals up to 100 MHz over the entire section is recommended. Following the normative specifications, however, the respective end plug is excluded from the calculation during a channel measurement by the measuring device and the use of appropriate channel adaptors on the measuring device – ‘connector compensation’. Since the end plugs can represent significant sources of error, especially in the case of field-configured cables with no intermediate patch bays/transition points, these should be taken into account in the measurement.
Tip: Depending on the measuring device used, this checks the end plugs at the beginning of the measurement and only masks them out, or continues with the channel measurement if the characteristics of the end plugs are sufficient for the performance class.
Less recommendable, but possible as an alternative, is the use of permanent link measuring adaptors. In the case of a permanent link measurement, the end plug connections are included in accordance with the standards; 2 double couplers are then required for the connection of the cable section/DuT.
If the combination of permanent link adaptor and channel limit values is not possible in your measuring device, the permanent link limit values EN50173-1, Appendix B Class D for signals up to 100 MHz can be used as an alternative. Since permanent link limit values are several dB ‘stricter’ than channel limit values, the channel limit values are usually complied with if the test according to permanent link limit values is passed, despite the use of 2 double couplers.
If Ethernet double couplers have to be used when using permanent link adaptors on the measuring device, high-quality all-metal executions must be used (at least EN50173 Cat. 6), in order to be as transparent as possible for the measurement.
Patch cable adaptors have been available for some certifiers since around 2012. These allow short point-to-point connections (“patch cables”) to be certified directly without masking out the end plugs. The TIA/SIO patch cord limit values must then be used in the measuring device.
Please note: in the case of patch cord measurements, some measurements are only informative according to TIA/ISO and no longer contribute to the PASS/FAIL decision.
Accordingly M12-CAT5e adaptors can be used for the measurement of M12-M12 patch cables.
For certification, the measuring device employed should attain at least accuracy level III according to IEC61935-1.

Explanations concerning certifier measurements for twisted pair cables

For specifications of the electrical limit values, refer to Cable specifications.

Measurements of Ethernet cables must take place in accordance with prEN50346:2001.

	Set limit values/limit value data record A limit value data record consists of several different limit value curves, e.g. resistance, delay, NEXT, etc. Depending upon the set limit value data record the individual measurements are active a measurement serves only for informative purposes (e.g. the length measurement in the EN50173 channel specification) or must be fulfilled (e.g. the attenuation measurement of the EN50173 channel specification) Each data record will trigger complex measurements and thereby produce multicolored diagrams; it is the user’s duty to select the correct limit values for its specific case.

	Frequency dependence of the parameters Most of the parameters shown below are measured over a given frequency range. In evaluating the result curves f(f), it must be taken into account that FastEthernet does not work at a constant frequency, see Basic Ethernet principles.

Measurement	Explanations
Wiremap	Continuity test of all connected cores 1-8, screen If, for example, a 4-core cable is measured, but an 8-core cable is specified in the device, the wiremap test and thus all subsequent tests will fail.
Resistance	DC resistance/loop resistance, given in Ω/100 m normal: 12 Ω/100m @AWG22, 19 Ω/100m @AWG26
Length	Normally measured via NVP, which must therefore be entered correctly in the cable data for the test. NVP (Normal Velocity of Propagation): ratio of the signal propagation speed in the cable to the speed of light; usually 60 – 80% and to be taken from the cable data sheet. Mainly results from the ‘length of lay’ and degree of twisting, e.g. where 2 m of Ethernet cable contain 2+x m stranded wires per core. The larger the NVP value, the less the cable elements are twisted. The cable length as such is not actually a critical value according to EN50173, but leads via length-dependent characteristic values (such as attenuation) to electrical problems or via propagation-delay-dependent processes to protocol problems.
Propagation Delay	Results from the propagation delay of the signal in the cable. Leads to problems if a permanent link measurement (specified at max. 90 m » usually 498 ns) is to be performed on a 100 m Ethernet cable.
Differential signal delay	Time delay in the signal propagation delay of a core pair. Should be 0 ns if possible.
Insertion Loss Attenuation	The parameter for the evaluation of the cable characteristics: the attenuation reduces the signal amplitude per meter of cable the attenuation is given as a positive value in [dB/100m] – SMALLER values are better here the attenuation is frequency-dependent: the higher the frequency, the higher the (real) attenuation is in the cable. As a result, the originally square signal from the transmitter is smoothed to the well-known ‘eye’ shape – the receiver must recover the signal by the use of equalizers an attenuation of 3 dB corresponds to a power loss of approx. 50% the attenuation increases - if the cable becomes thinner (AWG number increases) - if the cable is shielded (parasitic capacitances) - if stranded cores are used instead of rigid cores EN50173 permits different attenuation classes depending on the purpose (permanently installed or device connection = patch cable), see Limit value records. For orientation (according to EN50288-2:2003) Permanently installed cable: 21.3 dB/100m @ 100MHz (cable in stranded wire execution for moving operation is also available according to this specification!) Patch cable/device connection: 32 dB/100m @ 100MHz CAUTION: these are not the limit values according to which a complete cable section is specified in accordance with EN50173!
Return Loss	Waves transmitted into the cable are partly reflected back to the transmitter by defects. Defects may be in the material or at the plug transitions. The return loss is the difference between the signal transmitted into the cable and the signal reflected back. The higher the measured return loss the better; the attenuation is high and the value (re-)received signal is thus smaller Order of magnitude: 10 dB/100m @ 100 MHz for the EN50173 Channel Class D measurement.
NEXT PS NEXT	NEXT (Near End Cross Talk) describes the extent of the crosstalk from one pair of cores to a neighboring pair. For the measurement, a signal of a known strength is transmitted via pair X and the irradiation is measured on all neighboring pairs. 6 combinations are thus possible with 4 pairs. NEXT is measured at both ends (NEXT, FEXT); hence, 12 result curves f(f) are determined. NEXT has a negative measured value in [dB/100m]: the value describes the ‘volume’ of the received signal on the neighboring pairs of conductors in relation to the transmitted power – the more negative the value, the better. NEXT is usually applied positively without a sign for illustration purposes; positive LARGER values are then better than smaller ones. The longer a cable is, the more sensitive it is to NEXT. Since good twisting protects against NEXT, plug connections are particularly critical: a few mm of untwisted core pair in a plug significantly affect the measured value. Note: one plug/socket transition already creates an untwisted section of 1-2 cm! NEXT depends on the orientation of the cable section: a coiled cable will deliver a different NEXT result to an uncoiled and stretched cable. It is therefore preferable to carry out the NEXT measurement on the completed installation. PSNEXT (PowerSum NEXT) is calculated for each pair of cores as the sum of the crosstalk from all other pairs. If PSNEXT is also applied positively for illustration purposes, then LARGER values are better. the PSNEXT curves are typically a few dB worse than the NEXT results.
ACR-N ACR-F, ELFEXT PS ACR-N, PS ACR-F	ACR-N (Attenuation to Crosstalk Ratio, Near End) is calculated as the difference per cable pair between the worst results of the NEXT measurements and the attenuation measurements as a function of the frequency f(f). It therefore approximates to the worst signal-to-noise ratio and is thus an outstanding parameter with which to evaluate the quality of a transmission link. It is calculated for each core pair. The larger the value, the better – the receiver can then distinguish the wanted signal more clearly from interference. An ACR-N of 10 dB can be described as a well recognizable signal. ACR-F (Attenuation to Crosstalk Ratio, Far End) is subject to the length-dependent attenuation and is normalized from NEXT including the attenuation on length-independent values. It is also called ELFEXT (Equal Level Far End Crosstalk). The larger the value, the better. PS ACR is calculated as the difference between PS NEXT and the insertion loss and means the entire signal-to-noise ratio of a pair of cables. The larger the value, the better.