BUILDING A LOW COST ARTIFICIAL NETWORK FOR EMI ASSAYS
Abstract – In this paper will be presented a methodology for implementation of a Line Impedance Stabilization Network - LISN, symmetric commutable, in accordance with the specifications of the Normative IEC CISPR 16-1, using easily acquirable components in the electro-electronic stores. The Line Impedance Stabilization Network is used for conducted EMI assays in equipment’s witch current is not above 16 A.
1Introduction
To perform a assay of Conducted [1] Electromagnetic Interference (EMI) in laboratory it should own, besides an appropriate assay area and climate conditions, equipment’s that follows the IEC CISPR 16-1 [2] regulations. The principal equipment’s, indispensable for performing the assays, are the Artificial Network and the EMI Receptor.
In Conducted EMI assays, whose are detailed at [3], [4] and [5] references, the objective is to analyze the interfering voltage, generated by a distinguished equipment that allows to compare with the determined limits, resolved by a specific regulation, as, for example, The IEC CISPR 14 [6] Regulation. The artificial network is necessary to allows that the interfered signal be applied at the known impedance in the input power supply terminal of the under assay equipment, isolating the assaying circuit from the interferences provided by the power supply network and to couple the interferent voltage to the EMI receptor.
In Artificial Networks, the impedance measured between each terminal of the assayed equipment and ground must keep invariable, independent from the charge connected in their terminals, even when it is a short circuit, with a measure receptor connected or a equivalent resistance.
The presented Artificial Network is a commutable symmetric kind, it means, the user can perform measures as well as in the phase or in the neutral, just changing the key position. The presented example have as his most significant characteristic a 50 impedance associated in parallel with a 50 H inductor associated in series with a 5 resistance.
2Building an Artificial Network.
To build an artificial Network as the description presented in this paper, results in a significant economy in the implementation of a EMI implementation Laboratory, because in the end we will achieve an Artificial Network for Conducted EMI Assays with a 20% lower cost from a commercial one.
Figure 1 – Project of an Artificial Network
The project of this Artificial Network presented on Figure 1 – Project of an Artificial Network and its components are listed below on Table 1 .
COMPONENTS / VALUER1 / 5
R2 / 10
R3 / 1 000
R4 / 50
R5 / 50 (measure instrument impedance)
C1 / 8 F
C2 / 4 F
C3 / 0,25 F
L1 / 50 F
L2 / 250 F
Table 1 – List of components to build the Artificial Network.
The Artificial Networks impedance is defined by the components: L1, C1, R1, R4 and R5. And the components L2, C2 and R2 come power supply’s spurious isolated, being optional when the power supply’s network from the measure local is well fitted.
To the bigger capacitors, it was used alternated current capacitors mostly adopted in motors (polypropylene) achieving the low cost budget.
3Building the Inductor.
The L1 inductor is a coil with 35 espires, shaping one only layer of 6mm enameled wire. The step of this coil is 8 mm, rolled in an isolating core of 130 mm or 5 inches as the IEC CISPR 16-1 [2] regulations indicates.
The wire diameter is the dimension that considers to minimize the inductors resistive component. However, the coil built was fashioned using a 4 mm wire, because the current from the assays equipment’s is under 5A.
The core was fashioned with a 150 mm commercial PVC pipe with 280 mm length. The reduction of this diameter was obtained with a longitudinal slice taking off a 63 mm band. Inside the core, it was placed a 75 mm tube and the space between this tube and the 150 mm pipe was filled with expanded polyurethane to achieve a good mechanical resistence.
Figure 1 – Detail from the L1 Inductors fashion.
The step rolling control of the inductor was made using a 4 mm fishing string (nylon) between each espire.
To suppress internal resonance in this inductor the IEC CISPR 16-1 [2] regulation, establishes that 430 10% being connected between the espires 4 and 8, 12 and 16, 20 and 24 and 26 and 32 as showed in Figure 3 and in detail in Figure 4.
To have a 430 resistance, which is not commercial it was associated in parallel a 470 resistor with another of 4700 .
Figure 2 – Project of Resistors Positions in the Inductor.
Figure 3 – Detail of Resistors Positions in the coil.Detalhe da colocação dos resistores.
To hold the resistors to the coil, the enameled wire was scraped on the espires indicated by the scheme in Figure 3. After welding it was applied a thermal glue to achieve mechanical stability to these components.
4The Rack.
The inductor and the others components must be assembled inside a metallic rack. The base and the sides may be punched to allow heat dissipation when it is necessary. The suggested dimensions by the Regulation are 360 x 300 x 180 mm, Figure 5. Here, it was used a microcomputer rack, with dimensions 380 x320 x 180 mm, as showed in Figure 6.
Figure 4 – Rack Suggested by Regulation.
Figure 5 – Artificial Network Assembled on a Microcomputer Rack.
5Artificial Network Validation.
It is in the Regulation that the Artificial Network should have an equivalent impedance presented in Figure 7, with a 20% tolerance.
Figure 6 - Impedance of Artificial Network as agreeing with Regulation.
The validation of the assembled Artificial Network, was made with aide of a signal generator and a measuring signal levels instrument, besides a pattern terminal of 50 .
To execute the validation, the signal generator was adjusted to 9 kHz and the attenuation, caused by the Artificial Network, was measured. The process was repeated in others frequencies, evaluating the frequency band that the equipment will be used. With the results, a graphic was built as showed in the figure 8.
Figure 7 – Attest of Artificial Network Built.
The results obtained with the Artificial Network show that the impedance is out of the values specified by the regulation to frequencies under 20 kHz. These results do not make not viable, because the equipment that are going to be assayed, usually, works with frequencies above 20 kHz, where the Artificial Network achieve its objectives.
6Conclusions.
The mainly point of this work is in the fact of making more accessible the assays realization, diminishing the assembling costs in EMI assaying laboratories, reflecting even by academic or industrial environment.
The components utilized to build the Artificial Network are easily acquired, because they are not specific for this usage.
The only problem to use Artificial Networks is the calibration factor. However it is also necessary, for commercial equipment’s, to guarantee that the measures are correct, it is essential to track the equipment’s to known patterns internationally recognized.
7BIBLIOGRAPHIC REFERENCES.
[1]IEC 61000-1-1, Electromagnetic compatibility (EMC) – Application and interpretation of fundamental definitions and terms, 1 ed., Apr. 1992.
[2]IEC CISPR 16-1, Specification for radio disturbance and immunity measuring apparatus and methods – Part 1: Radio disturbance and immunity measuring apparatus, 2 ed., Oct.1999.
[3]MAGNUS, ELIO F., 2001, Desenvolvimento de uma ferramenta para ensaios de EMI Conduzida de baixo custo. Dissertação de M.Sc., Pontifícia Universidade Católica do Rio Grande do Sul, Brasil.
[4]DOS REIS, F. S., 1995, Estudio y Criterios de Minimizacion y Evaluacion de las Interferencias Eletromagneticas conducidas en los convertidores ca – cc. Tese de D.Sc., Universidade Politécnica de Madri, Espanha.
[5]IEC CISPR 16-2, Specification for radio disturbance and immunity measuring apparatus and methods – Part 2: Methods of measurement of disturbances and immunity, 1.1 ed., Aug.1999.
[6]IEC CISPR 14, Limits and methods of measurement of radio disturbance characteristics of electric motor-operated and thermal appliances for household and similar purpouses, electric tools and electric apparatus, 3 ed - amendment 1, Aug.1996.