mopdf.com

2016 IEEE International Symposium on Electromagnetic CompatibilityHardware Demonstration:Radiated Emissions as a Function ofCommon Mode CurrentJuly 26, 2016John McCloskeyJen RobertsNASA/GSFCChief EMC [email protected]&D Inc. work performed for NASA/GSFCEMC [email protected] be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.1

Acronyms Common mode conducted emissions (CMCE) Conducted emissions (CE) Electric field per unit current (E/I) Electromagnetic interference (EMI) Equipment under test (EUT) Radiated emissions (RE)To be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.2

Introduction (1 of 3) Due to the time- and resource-consuming nature of the test for radiatedemissions, electric field (RE02/RE102), it is often prohibitive to performearly diagnostics by performing the test or even an abbreviated version of it Fortunately, it is possible to perform a much simpler, cheaper test – acommon mode conducted emissions (CMCE) test – which will give usefulpredictions for much less expenditure of time and money The CMCE measurement can be easily performed in the hardwaredevelopment lab in order to provide an early indication of whether theEquipment Under Test (EUT) will pass the RE102 test before the EUT istaken to a full EMI test facilityTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.3

Introduction (2 of 3) At frequencies below 200 MHz, a significant portion of the radiated energyoriginates from uncontrolled common mode currents on cables connectedto the unit In this demonstration, a controlled current is applied to a 1 m wire 5 cmabove a ground plane, and the resulting electric field is measuredRF AMPLIFIER(as needed)INIOUTRF SIGNALGENERATORr 1m80-90 cmEMEASUREMENTANTENNA120 cmPREAMPLIFIER(as needed)INOUTSPECTRUMANALYZERINTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.4

Introduction (3 of 3) The transfer function of electric field per unit current (E/I) is determined Presented as a tool for predicting radiated electric fields from a simple measurementwith a clamp-on current probe before the product ever leaves the developmentlaboratory E/I correspondence is evaluated at frequencies 30 MHz Per MIL-STD-461, RE102 below 30 MHz is measured with rod antenna, whichresponds to potential, not current Product development engineers are encouraged to perform these measurements inorder to facilitate diagnosis of potential problems as early as possible in theproduct's development cycleCURRENT PROBE PLACED ATVARIOUS LOCATIONS ALONG WIRERF AMPLIFIER(as needed)INOUTRF SIGNALGENERATORPREAMPLIFIER(as needed)INOUTSPECTRUMANALYZERINTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.5

First, a bit of theory.Electric Field fromElemental Electrical (Hertzian) DipoleElectric(Hertzian) Dipolez Hφθr Eθdlx ErEφ 0 11 jβ 0 rIdl2 Er 2 η 0 β 0 cos θ 2 2 j 3 3 e4πβ0 r β0 ryEθ φWavenumber:β0 2πλ 2πfc 111 Idl 2 2 j 3 3 e jβ 0 rη 0 β 02 sin θ j4πβ0 r β0r β0 rTHEORYEXPERIMENTTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.6

A Very Sick Integral For a wire of any finite length, a precise calculation of the electric field at distance rwould require a very complex integral Distance to measurement point must be varied with location of dl along the wire Relative phase of each field contribution must be considered Vector contributions of Er and Eθ Etc., etc., etc.lRF AMPLIFIER(as needed)INIOUTRF SIGNALGENERATORr 1mEMEASUREMENTANTENNA120 cmPREAMPLIFIER(as needed)INOUTSPECTRUMANALYZERINTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.7

Simplifying Assumptions for E/I Envelope Measurement is in traditional definition of “far field” βr 1 (f 48 MHz for r 1 m; generally OK for f 30 MHz ) 1/βr term dominates 1/(βr)2 and 1/(βr)3 terms may be neglected Cable behaves as a point source for estimating worst-case envelope All current carrying elements are assumed to be at the same distance r fromthe measurement point The electric field contributions from all current-carrying elements areassumed to be in phase at the measurement pointEmax µ 0lf I2rTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.8

Transmission Line Current Model Transmission line current has horizontal and vertical components Resulting electric field will have horizontal and vertical componentsV(z)I(z)LΔzdId ΔzV(z Δz)I(z Δz)CΔzTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.9

E/I Envelope for Wire 5 cm Above Ground Plane Math in backup charts, but here's the bottom line: Horizontal polarization For cables of l 1 m (typical on most spacecraft), E/I is essentially independent of wire length above 48 MHz (βl 1) For most spacecraft, RE102 is mainly a concern for f 200 MHz High frequency asymptote determined largely by interaction of cable with ground plane (details in backup slides)Vertical polarization E/I is essentially independent of wire length for f 30 MHz Equivalent to E/I for horizontal polarization for wire of l 1 m High frequency asymptote determined by monopole formed between cable and ground plane (details in backup slides)To be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.10

E/I Measured vs. PredictedTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.11

Current Measurement Helpful Hint For f 30 MHz, cables of l 1 m will exhibit standing waves Current will NOT be constant along length Note that E/I is based on integrated, i.e. average, current along cable Traditional approach Place spectrum analyzer in max hold mode Physically scan probe along length of cable Relatively quick and easy, but this will give the peak current on cable, notaverage Will overestimate current and electric field Recommended approach Measure current at many locations along wire Take average If measuring in dB, convert to numeric firstTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.12

Summary Radiated emissions and conducted emissions "joined at the hip"RECE E/I transfer function envelope provides useful tool for predicting radiatedemissions early in product development cycle The current probe is your friend; measure those common mode currentsearly and often Easy and useful measurement to make in hardware development lab beforetaking product to EMI facility Don't have to fight RF background, room resonances, etc. Identify specific sources of potential problems as early as possible Control those common mode currents Provide the desired low impedance path (e.g. shield, ground plane, etc.) andmake them flow where you want them to Let nature do the work When you control common mode currents, you go a long way towardcontrolling radiated emissions and most other EMI problemsTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.13

BackupTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.14

Electric Field as Function of Current(Electrically Short Cable, Hertzian Dipole Model)z Hφθr Eθdlx ErEθdl η 0 β 024πmaxIyEθmaxIβ0 2πλ 2πf 2πf µ 0ε 0c2 1 11 2 2 3 3 β0 r β0r β0 r 2Far field approximation (β0r 1):φBut: 111 Idl 2 2 j 3 3 e jβ 0 rη 0 β 02 sin θ j4πβ0 r β0r β0 rEθ Electric(Hertzian) Dipoleand η 0 EθmaxIEθmaxI l η 0 β 024πµ0ε02 1 l η0 β 04πr β0r η 0 β 0 2πf µ 0ε 0 µ0 2πfµ 0ε0l2πfµ 04πrµ 0l f2rTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.15

Transmission Line CurrentsLΔzV(z)I(z)dId ΔzV(z Δz)I(z Δz)Vertical:l1 lsin βlI0 cossinββzIIzdz00βlβll 00βl 1:I(z) I0cosβzCΔzHorizontal:IH I(z) I0e –j βzIHI0 1ENVdI d ( z ) z I ( z ) I ( z z )dI d (z ) I (z ) I (z z ) dI (z ) βI 0 sin βz zdzI V β I 0 sin β zdz I 0 cos β z 0 I 0 (cos β l 1)ll0βl 1:IHI0 ENV1βlIVI0 2ENVTo be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.16

Horizontal (Wire) CurrentIH/I0 (Dimensionless)2l l l l l l l l 25 cm envelope25 cm50 cm envelope50 cm1 m envelope1m2 m envelope2m10101001000Frequency (MHz)To be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.1000017

Vertical (Displacement) Current5Envelopel 2ml 1m4l 50 cmIV/I0 (Dimensionless)l 25 cm321010100Frequency (MHz)To be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.100018

Response of sin(x)/x :0 for x nπ1/x for x (n 1)π/2Envelope of sin(x)/x :1 for x 11/x for x 1 sin(x)/x (linear)Response and Envelope of sin(x)/xEnvelope11/x sin(x)/x 00123π45x6782π9103π sin(x)/x (dB)10Lines meetat x 10 dB/decade0-1020*log10 sin(x)/x -20-30-400.11x10To be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.19

Horizontal E-Field (No Ground Plane Attenuation)50l 2 m envelopel 1 m envelopel 50 cm envelope40EWI0l 25 cm envelopeEW/I0 (dBΩ/m)EWI0 ENV ENVµ0lf2r µ lf111 0 βl2r 2πlf µ0ε 0 4πrµ0 120π ε04πr30 30 Ω / m 30 dBΩ / m(1m)3020EWI010EWI04π 10 7lf 4π 10 7 lf 2(1m)(ENV) 20 log10 f MHz 20 log10 l 4dBENV , dB0101001000Frequency (MHz)To be presented by John McCloskey and video recorded at the 2016 IEEE International Symposium on Electromagnetic Compatibility, Ottawa, Canada, July 26, 2016.1000020