Localization and Identifying EMC interference Sources of a Microwave Transmission Module Ph. Descamps 1, G. Ngamani-Njomkoue 2, D. Pasquet 1, C. Tolant 2, D. Lesénéchal 1 and P. Eudeline 2 1 LaMIPS, Laboratoire commun CRISMAT, UMR 6508 CNRS, 6 bd Maréchal Juin, Caen, France. 2 Thales Air Systems SA, Technical Unit Radio-Frequency (TU-RF), Technology and Innovation (REIRI-Y), Z.I. du Mont Jarret, 76520 Ymare, France June 2012
OUTLINE Motivation Description of the Demonstrator Expression of the electromagnetic fields in a cavity 3D Electromagnetic Simulations of the demonstrator Electric field distribution propagation of the energy in the cavity S parameters measurements Electrical field measurements with GTEM cell Conclusion
BACKGROUND French national projet : AUDACE (Analyse des causes de DéfaillAnces des Composants des systèmes Electroniques embarqués) Part of AUDACE project: Measuring stress and their effects Aim of this work : Reliability of components and circuits against electromagnetic disturbances : Consideration of EMC in the realization of transceiver modules. digital processing Receiver Microwave stage (LNA..) Receiver duplexer Management of radar Signal generator Transmitter Synoptic of the radar Transmitter Antenna Identify problems of EMC identify sources of disturbance.
Presence of different signals in the same transmission module Motivation Transmission/receivers (T/R) are usually shielded by metal cavities to be immune from external disturbances and to avoid disturbing close circuits. Very high radiated power emission due to high output power of transmitters (radars) (hundred of kilowatts). Transmitters modules are built in a confined area containing several radiating elements (ex : inductances, transmissions lines active circuits..) Measuring of the electromagnetic field radiated Identifying critical high frequencies : parasitic frequency noise and resonance frequencies Identifying areas of strong field and components generating high emissions. Studying the effect of the metal shielding on the electromagnetic field. Complete transmission module too complex to study in terms of EMC Module simplified reproducing the main effects of EM
Description of the Demonstrator Circuit representation of the simplified demonstrator A demonstrator in case of reproducing the maximum electromagnetic effects found on the circuit : measurement of different signals as microwave signals, power, digital, low frequency signals. Patch antenna Microwave Feeder coupler inductance Using a circuit bulk 5 Layout of the PCB The microstrip line is a 150 mm-long Zc = 50 Ω Coupler = 20 db at 3 GHz The two antennas have been calculated for a 5 GHz resonance frequency Feed line w = 3mm, l = 7.07mm 1 square patch antenna and 1 patch antenna with notches Length of the package: 150mm, Width : 55mm, Height : 75mm Substrate : FR4, h = 1,6mm and εr = 4,6 The complete demonstrator has been simulated and measured as a 6-port with and without the shielding box.
Electromagnetic fields in a cavity Cross section f m,n,p 2 C r m a 2 n b 2 p c 2 Photography of the cavity The calculation of cutoff frequencies of the modes allows us to identify each mode and we can then select the TE mode (Transverse Electric) or TM mode (Transverse Magnetic) using Mapping of electromagnetic fields.
3D Electromagnetic Simulations of the demonstrator Cartography of the electric field for a half-filled cavity with air. Dimensions : 55 x 37,5 x 150 mm Two degenerate modes observed correspond to the modes TE11 and TM11. (confirmed by dispersion diagram).
Propagation of electromagnetic fields Propagation of electromagnetic fields of the fundamental mode for F = 3GHz E field Mode TE 10 H field Mode TE 10 Over the cutoff frequency of the fundamental mode F = 3GHz, the electric field propagates along the cavity. Distribution of electromagnetic fields for F = 1GHz E field Mode TE10 H field Mode TE 10 At 1 GHz, the electric field does not propagate in the guide: it is rapidly attenuated.
Propagation of the electric field of higher modes at F=3GHz The higher modes are propagated in the cavity when the frequency is higher than cutoff frequency There is a stationary phenomenon below the cutoff frequency and propagation above the cutoff frequency. Propagation of the electric field of higher modes for F = 3GHz
Identification of index modes The resonance frequencies have been calculated, simulated and measured Cutoff frequencies for the half empty box Mode TE 101 TE 102 TE 103 TE 011 TE 012 TM 110 TE 104 TM 111 TE 111 Calculated 2.90 3.38 4.05 4.12 4.47 4.84 4.84 4.94 4.94 Calculated with correction 2.85 3.32 3.98 4.05 4.39 4.76 4.76 4.86 4.86 Simulated (3D simulator) 2.85 3.32 3.97 4.12 4.46 4.74 4.83 4.88 4.93 Frequencies of the resonance modes in GHz (1 st line calculated for the empty box, 2 nd line calculated with the correction of εrequ, 3 rd line simulated with substrate and without printed metal, 4 th line measured on S 41 ) Cross section a = 5.5 cm ; b = 3.75 cm ; c = 15 cm f m,n 2 C r m a 2 n b 2 (Correction : ) 10
Electric field distribution propagation of the energy in the cavity Electric field distribution Propagation of the electric field along the line at F = 3GHz. (Vector representation) Propagation of the energy in the cavity (F = 3GHz) Substrate FR4 : permittivity 4,5 ; height 1,6mm The energy is spread in the substrate and symmetrically with respect to the axis line
Measurements of the coupler : S31 and S41 Mode TE 101 TE 102 TE 103 TE 011 TE 012 TM 110 TE 104 TM 111 TE 111 Calculated 2.90 3.38 4.05 4.12 4.47 4.84 4.84 4.94 4.94 Calculated with correction 2.85 3.32 3.98 4.05 4.39 4.76 4.76 4.86 4.86 All Simulated the S-parameters (3D simulator) 2.85 3.32 3.97 4.12 4.46 4.74 4.83 4.88 4.93 between SMA ports were extracted with open and closed box (with Measured (S parameters) 2.85 3.32 3.85 3.95 4.33 4.67 4.72 4.8 4.83 shielding) Coupling : C = 20 log(s31) = 20dB at 3GHz Isolating : I = 20 log (S41) = 30dB at 3GHz Directivity : D = 10dB Parasitic modes appear as perturbations for the measured S31 and S41 between the line and the coupled access of the coupler. Resonance frequencies have been identified in the table below :
Measurement of the square patch antenna : S51 Return loss S55 of the square patch antenna The same parasitic modes appear with S51 as perturbations between the line and the square patch antenna.
Measurement of the square patch antenna : S 61 Return loss S66 of the square patch antenna Measured resonance frequencies are very close to calculated and simulated resonance modes of the cavity. Resonance frequencies appear only when the cavity is closed.
Comparison of S parameters of the circuit with opened and closed shielded box. All perturbations find by the resonance frequencies when the cavity is closed are identified as parasitic modes.
1,04 m Description of GTEM cell for electric field measurements The GTEM cell is a frequency extended variant of the traditional TEM (Transverse Electro-Magnetic) cell. The GTEM cell is, in principle, a tapered coaxial line (offset septum plate), from a coaxial feeding point, having an air dielectric and a characteristic impedance of Zc = 50 Ω. This coaxial line is terminated by a combination of discrete resistors and RF absorbers to achieve a broadband match. The outer conductor of this coax line is created by the metal walls of the cell which provide screening for both internal and external electromagnetic fields. Typical Test Set-up for RF Emission Specifications Septum height: 500 mm Dimension (LxWxH in m): 2.95 x 1.48 x 1.61 Door (LxH in m): 0.44 x 0.38 EUT max. size (LxWxH in m): 0.41 x 0.41 x 0.31 EUT size (3 db criteria, LxWxH in m): 0.30 x 0.30 x 0.15 Max input power: 100 W RF-input connector: N-type Nominal impedance: 50 Ω Frequency range: DC up to 20 GHz Test Cells for EMC Radiated & Immunity Testing DC to 20GHz Volume for testing 16
Electric field measured in GTEM cell The closed demonstrator has been put into a GTEM cell. The input port 1 has been fed with a 20dBm RF signal. The amplitude of the electric field has been measured by the septum in the three XYZ directions as defined in Figure below : E Ex 2 Ey 2 Ez 2 Electric field component Ez is higher than the two other Electric field components (Ex and Ey).
Comparison between the S 41 of the coupler and the measurement of the electric field. The same parasitic modes appear as perturbations between the line and the coupled access of the coupler with the measurement of the electrical field.
Comparison between the S 51 and S 61 between the line and the two square patch antennas and the measurement of the electric field. Other parasitic modes appear as perturbations between the line and the two square patch antennas with the measurement of the electrical field. Measured resonance frequencies are very close to the calculated and simulated resonance modes. Mode TE 101 TE 102 TE 103 TE 011 TE 012 TM 110 TE 104 TM 111 TE 111 Calculated 2.90 3.38 4.05 4.12 4.47 4.84 4.84 4.94 4.94 Calculated with correction 2.85 3.32 3.98 4.05 4.39 4.76 4.76 4.86 4.86 Simulated (3D simulator) 2.85 3.32 3.97 4.12 4.46 4.74 4.83 4.88 4.93 PIERS 2011, Marrakech Measured (S parameters) 2.85 3.32 3.85 3.95 4.33 4.67 4.72 4.8 4.83
Comparison between the S41 of the coupler and the measurement of the electric field. The same parasitic modes appear as perturbations between the line and the coupled access of the coupler with the measurement of the electrical field.
Comparison between S parameters of the coupler, two patch antennas and the measurement of the electric field By measuring the electric field in the GTEM cell, it is possible to identify parasitic frequencies (modes) sources of perturbation
Correlation between electric field orientation (xoz plane) and S41, S51 and S61. The orientation of the electric field in xoz plane is more significant (0 corresponds to Oz axis) than in the other planes. Peaks appear for the identified resonance frequencies.
Parasitic Electric field : Contribution of cables 55dB The contribution of the radiated electrical field by cables connected to the module (empty cavity + circuit) is negligible.
Conclusion & Perspective Electromagnetic behavior of a demonstrator including several elements in case of reproducing the maximum electromagnetic effects found on the circuit has been studied. Parasitic resonances can be detected from measurements of the electric field outside the closed box identify sources of disturbances. As the field patterns of all the modes are known, it is thus possible to know what the field repartition inside the box is and to know where it is adequate not to put components that are liable to radiate. Better understanding the interaction between the cavity and high frequency circuits. Appropriate probes to scan the surface of the circuits can be built to measure electromagnetic near fields and to validate the method.
THANK FOR YOUR ATTENTION