Conducted Emissions measurements set up
At lower frequencies emission of RF power from EUT is presumed to take place by the cabling of an equipment principally. Within the framework of the CE mark related EMC testing considered frequency range is 150 kHz to 30 MHz. Levels of emission are expressed in terms of the log of micro volts (dBµV). Common limit levels are starting at 46 dBµV.
LISN / AMN
Traditionally the emission of the mains cables has been measured using a AMN (Artificial Mains Network) or LISN ( Line Impedance Stabilization network). This network shows a stabile and defined load impedance to the EUT. The applied network is of the V-type, this means that both symmetrical and non-symmetrical interference contribute to the measured value. The LCL value of the standard V-network is -6 dB. Symmetrical interference therefore is attenuated 6 dB before is is lead to the measuring receiver. Common mode voltages are not attenuated.
Definition of terms:
Symmetrical / asymmetrical interference
Symmetric signals in a set of conductors are identical in value but of same frequency and opposite phase. Summing all components of a symmetric signal results in zero. Asymmetric interference is a signal that has no counterpart in other conductors (within the same cable). Summing all interference currents in a set of conductors (cable) results in a non-zero component, the asymmetric interference component or Common Mode current.
Symmetric interference is normally coped with by short circuiting conductors at the exit port of an EUT. Asymmetric components are normally attenuated by inserting impedance. These simple solutions works only if the signal frequencies are essentially different from the interference frequencies such as in a mains supply cable, we speak about "out of band" interference.
Telecommunications related networks
The problem of modern telecommunication cables is that they contain signal frequencies within the same frequency range as expected and to be tested interference signals (=in-band interference). Therefore specific networks have been developed that cope with the interference characteristics of cable networks. Telecommunication cable do not only emit CM spurious signal coming out of the EUT, but also part of their data signal content, caused by cable imperfections (both by design as by construction imperfections). These unwanted conversion of signal current into interference current in cabling is being characterized by a term called Longitudinal Conversion Loss. Telecom ISN networks are used to mimic the behavior of a telecommunications cable and allow for connecting a measurement device. To mimic different kind of cabling quality they come in a number of LCL values.
The LCL value defines the amount of DM signal (symmetric) that is being converted to CM signal (asymmetric) in the network. The LCL factor therefore is a measure for the expected cable quality. The selected ISN network should therefore match the expected LCL value of the network cable to be tested.
Normally the only phenomena causing spurious emission is a CM current on a cable. Traditionally this has been measured on the EUT exit port. However, part of the signal current is being transferred to CM current due to imperfections in the cable symmetry. (part of the signal energy is directly emitted by the cable (enclosed surface of cable segment partly compensated by twisting), this is not considered here). At lower frequencies it is not easy to add enough cable length to actually create enough cabling induced CM current at the cable exit port. For this a cable length up to ¼ lambda is required. If there were enough cable length, simply measuring CM current would have been sufficient. Normally, during test house emission measurement set up there is not enough cable length, and it is therefore that telecommunications conducted emission networks need LCL value; a purposely induced asymmetry mimicking the expected and prescribed cabling. A standard V-network (LISN) has a LCL value of -6 dB. For telecom ISN the value may vary between -20 and -80 dB corresponding to different categories of standard network cabling.
LCL = 0 dB equals that characteristics of infinitely bad cable pair
LCL = -60 dB represent a good quality twisted pair ( Cat 5)
Limits for conducted emissions
Currently the limits for conducted emissions are 56 dBµV (QP), with relaxation's up to 10 dB at the lower (<500 kHz) and 4 dB at the higher (>5 MHz) end of the measurement spectrum (residential equipment = Class B). In addition to the Quasi Peak requirements a Average measurement limit has been introduced at 10 dB lower level.
Some standards such as EN 55022 allow equipment for Industrial environments to comply with higher limits, but need to carry a warning label that the equipment may cause radio interference.
This warning label clause is subject of discussions between the European Commission and CENELEC, as the subject of conditional approvals is not within the mandate of CENELEC. The European Commission strongly discourages the application of Class A ITE (industrial emission levels) within residential areas.
Measurement accuracy for emission measurements
At this moment the accuracy of the EMC measurement need not be taken into account when obtaining the test result. It has been said that the original limits were defined taking into account the obtainable accuracy.
Reproducibility however, has to be maintained by selecting approved measuring instruments such as a CISP16 compliant receiver and verified test site.
The emission properties of an apparatus are established generally for one prototype only. Production of any electronics is subject to tolerances, in manufacturer and in component tolerances. This implies that the emission profile of any piece of electronics is subject to variations. The variations can be represented by a normal distribution. The CISPR limit demands that of the products put on the market at least 80% fulfills the emission limit requirement with a confidentially level of 80%. The consequence is that a prototype that fulfills the emission limit requirement does not guarantee any compliance confidentially. A one (1) sample test has 0% confidentially. A number of samples need to be evaluated in order to be able to declare compatibility. Many CISPR standards contain a paragraph showing calculation examples and recommended sample numbers to make such a statistical calculation.
Test set up
The test set up has been defined within the CISPR standards in CISPR 22 and in CISPR16. The EUT needs to be maintained at 1 meter from the AMN/LISN port. Excess cable needs to be folded in a non-inductive way in order to not influence measurements,
The edge of the EUT is kept at 40 cm from a vertical ground reference plane, so as to define the capacity to this wall in a reproducible way.
The LISN ground needs to be well connected to this ground plane.
At ce-test, conducted measurements are carried out in a shielded room, at 40 cm from one of the internal (conductive) wails. To comply with CISPR full compliance requirements the LISN output signals are measured using a R&S measuring receiver.
The equipment is under annual calibration and subject to a monthly verification schedule, to provide our customer with traceable and accurate test results.
Test set up example using a demo EUT.
This test setup allows us for testing EUT's up to 25 Amps at 230 Vac 3-phases, and provide customer with a full compliant test result.
Frequency Range 148.5 kHz - 30 MHz