Essential New Tools for EMC Diagnostics and Testing Todd H. Hubing Michelin Professor of Vehicular Electronics Clemson University
Where is Clemson University? Clemson, South Carolina, USA Santa Clara Valley Chapter Mtg. September 11, 2007 2
Santa Clara Valley Chapter Mtg. September 11, 2007 3
Santa Clara Valley Chapter Mtg. September 11, 2007 4
Santa Clara Valley Chapter Mtg. September 11, 2007 5
Tools Required to Outfit an EMC Laboratory Semi-anechoic test facility Shielded room Network analyzers Spectrum analyzers Impedance analyzers Waveform generators Oscilloscopes Antennas, Probes Power supplies Prototyping facilities Amplifiers, Pre-amplifiers Santa Clara Valley Chapter Mtg. September 11, 2007 6
Near-Field Scanners X,Y,Z Motion Positioner Motion Controller Measurement Equipment Probe DUT Santa Clara Valley Chapter Mtg. September 11, 2007 7
Near-Field Scanners Santa Clara Valley Chapter Mtg. September 11, 2007 8
Near-Field ESD Scanners ESD Scanner The system moves injection probes to predefined locations, injects pulses and observes the system response. In most cases, pulses are ESDlike, e.g., having rise times 0.1-2 ns. Injection is done using different injection probes for testing direct coupling, E and H-field coupling. To determine the voltage at an input of an IC at a level that leads to a mal-function, the voltages are measured. Santa Clara Valley Chapter Mtg. September 11, 2007 9
Near-Field ESD Scanners 310 Net 2 300 Net 4 290 450 280 400 270 350 260 Net 1 Net 3 300 Net 2 Net 4 250 250 240 180 190 200 210 220 230 240 250 50 100 150 200 250 Net 1 Net 3 Santa Clara Valley Chapter Mtg. September 11, 2007 10
Time/Frequency Domain Analyzers Source Identification Source Characterization Narrow-Band Transient Capture Santa Clara Valley Chapter Mtg. September 11, 2007 11
Time/Frequency Domain Analyzers 1400 1200 Frequency (MHz) 1000 800 600 400 200 27 28 29 30 31 32 33 34 35 36 Time (usec) 0.2 0.1 Voltage (volt) 0-0.1-0.2 27 28 29 30 31 32 33 34 35 36 Time (usec) Santa Clara Valley Chapter Mtg. September 11, 2007 12
Mini-TEM Cells DUT (IC chips) mounted at the inner side of the test board (within TEM cell) 50 Ohm load Santa Clara Valley Chapter Mtg. September 11, 2007 13
Mini-TEM Cells Standards SAE J1752/3 and IEC 61967-2 Santa Clara Valley Chapter Mtg. September 11, 2007 14
Mini-TEM Cells More attention is being focused on the EMC design of ICs, MCMs and other electrically small sources. These devices generally don t radiate directly themselves. They couple to larger objects (e.g. cables and enclosures) that serve as the antennas. Santa Clara Valley Chapter Mtg. September 11, 2007 15
Mini-TEM Cells There are only three possible coupling mechanisms from an electrically small source. Conducted Electric-Field Coupling Magnetic-Field Coupling Santa Clara Valley Chapter Mtg. September 11, 2007 16
Mini-TEM Cells Electric field coupling can be represented with a mutual capacitance, C TEM. The voltage coupled to either end of the TEM cell will be identical. TEM (a) (b) Santa Clara Valley Chapter Mtg. September 11, 2007 17
Mini-TEM Cells Magnetic field coupling can be represented with a mutual inductance, M TEM. Voltage appears across both terminations with opposite phase. TEM (a) (b) Santa Clara Valley Chapter Mtg. September 11, 2007 18
Mini-TEM Cells A hybrid can be used to differentiate electric and magnetic field coupling. The A-B output indicates the amount of magnetic field coupling. The A+B output indicates the amount of electric field coupling. Connected to Network Analyzer Port 1 Connected to Network Analyzer Port 2 A 0º TEM Cell Connected to Network Analyzer Port 2 Hybrid 0º 180º 0º C=A+B D=A-B B Santa Clara Valley Chapter Mtg. September 11, 2007 19
Voltage-Driven Radiation Mechanism C trace I CM V DM C DM I CM V CM C board C cable Far-Field Radiation V CM C C trace board V DM Far-Field Radiation V V DM V V CM Ground Ground Santa Clara Valley Chapter Mtg. September 11, 2007 20
Correlating C trace to C TEM C trace TEM V DM C DM I CM C board V CM C cable C trace C TEM /2.1 Santa Clara Valley Chapter Mtg. September 11, 2007 21
Effect of Extended Cable on Ground 1m l Santa Clara Valley Chapter Mtg. September 11, 2007 22
Mini-TEM Cells Current-Driven Common-Mode (Magnetic-Field) Coupling L trace I CM V DM I DM L ret Z L I CM V ret C ant Source can be fully characterized by the current I DM and the mutual inductance (source loop to antenna loop). Santa Clara Valley Chapter Mtg. September 11, 2007 23
Mini-TEM Cell By connecting both outputs of the TEM cell to a hybrid, it is possible to separate the electric field coupling from the magnetic field coupling. Magnetic-Field coupling is fully characterized by the source current and mutual inductance to the radiating structure. These are both determined by the TEM cell measurement. Electric-Field coupling is fully characterized by the source voltage and the capacitance of the device being driven to infinity. These can both be determined by the TEM cell measurement. Therefore, a TEM cell measurement can be used to extract the parameters required to predict maximum radiated emissions due to coupling from an electrically small source. Santa Clara Valley Chapter Mtg. September 11, 2007 24
Numerical Electromagnetic Modeling Tools Santa Clara Valley Chapter Mtg. September 11, 2007 25
Numerical Electromagnetic Modeling Tools COMPLIANCE GEMACS EZ-FDTD Q3D SuperNEC Microwave Studio MAGNETO FIDELITY Accufield COMSOL Multiphysics FEKO FLO/EMC EMAG SPEED2000 PAM-CEM MiniNEC SINGULA EMAP COULOMB Santa Clara Valley Chapter Mtg. September 11, 2007 26
http://www.cvel.clemson.edu/modeling/software/ Santa Clara Valley Chapter Mtg. September 11, 2007 27
EMC Expert Systems Printed Circuit Board Layout Automotive EMC System-Level Extensions Santa Clara Valley Chapter Mtg. September 11, 2007 28
Expert System Algorithms are constantly answering the question, How bad could it be? Santa Clara Valley Chapter Mtg. September 11, 2007 29
Effect of Extended Cable on Ground 1m l Santa Clara Valley Chapter Mtg. September 11, 2007 30
NCMS Board Analysis using Expert System Algorithms 8 clock buffers 28 load capacitors U2 U7 32 decoupling capacitors Clocked at 50 MHz No heatsink Size: 3 by 2, 6 layers Powered with one cable Santa Clara Valley Chapter Mtg. September 11, 2007 31
Measurement vs. Calculation: 1-nF Load Santa Clara Valley Chapter Mtg. September 11, 2007 32
PCB Expert System Structure Identify Sources Identify Antennas Evaluate Coupling Santa Clara Valley Chapter Mtg. September 11, 2007 33
PCB Expert System Structure Source Model Coupling Path Model Antenna Model Santa Clara Valley Chapter Mtg. September 11, 2007 34
PCB Expert System Emissions Models Differential-Mode Radiation Coupling to I/O Radiation Voltage-Driven Common-Mode Radiation Current-Driven Common-Mode Radiation Power Bus Radiation Santa Clara Valley Chapter Mtg. September 11, 2007 35
Current-Driven Radiation Model V 75 ωlpi, IDM, i, f MHz a ε r = 4.71 10 Lpi, IDM, i 75, f MHz a εr a εr ret, i 8 E cable-to-board 0.365 100 Vret 2 100 + 1 ( ωc ) D. M. Hockanson et. al., "Quantifying EMI resulting from finite-impedance reference planes," IEEE Trans. on EMC, vol. 39, no. 4, Nov. 1997, pp. 286-297. H. Shim et. al., Expert system algorithms for identifying radiated emission problems in printed circuit boards, Proc. of the 2004 IEEE International Symposium on EMC, Santa Clara, CA, USA, Aug. 2004, pp. 57-62. Santa Clara Valley Chapter Mtg. September 11, 2007 36 B 2
Power Bus Radiation Model y b M s Q( f) 1 1 1 = + + Q Q Q d c comp 1 z M s M s E = 120Ii h Q( f) ε min( ab, ) r r h M s a x H. Shim and T. Hubing, Estimating radiated emissions from the power planes in a populated printed circuit board, IEEE Trans. on Electromagnetic Compatibility, vol. 48, no. 1, Feb. 2006. Santa Clara Valley Chapter Mtg. September 11, 2007 37
PCB Expert System Structure Santa Clara Valley Chapter Mtg. September 11, 2007 38
System-Level Extensions Santa Clara Valley Chapter Mtg. September 11, 2007 39
System-Level Extensions Santa Clara Valley Chapter Mtg. September 11, 2007 40
PCB Expert System Algorithms The goal is to distinguish between a good design and a bad design and identify features of a design that are likely to result in emissions or susceptibility problems. Existing expert system tools are capable of finding many problems that would be difficult to locate manually. Santa Clara Valley Chapter Mtg. September 11, 2007 41
Validation of Software Santa Clara Valley Chapter Mtg. September 11, 2007 42
Validation of Software Santa Clara Valley Chapter Mtg. September 11, 2007 43
Validation of Software Santa Clara Valley Chapter Mtg. September 11, 2007 44
Validation of Software GRESET U4 DATA2 View of left half of board showing the problem nets. U6 U1 U20 CONNECTOR P2 GRESET CONNECTOR P1 DATA2 Santa Clara Valley Chapter Mtg. September 11, 2007 45
Validation of Software Santa Clara Valley Chapter Mtg. September 11, 2007 46
Validation of Software U4 Current-Driven Common-Mode Problem. U1 U37 $1I6\CLKCPU CONNECTOR P2 $1I6\CLKCPU CONNECTOR P1 U19 Santa Clara Valley Chapter Mtg. September 11, 2007 47
Maximum Allowable Power Bus Voltage E = 20 I V 37 ohm min ( max) f max ( θ, k) = 20 2. 76 0 cable _ rad _ 2πl sin factor 1.0 cable λ λ when lcable 4 otherwise 2π l λ sin board when lboard board _ size _ factor λ 4 1.0 otherwise E = E cable _ rad _ factor board _ size _ min factor Expert System Field Calculation Performed in Reverse E min 37 ohm V min = cable rad factor board size factor 20 2.76 Santa Clara Valley Chapter Mtg. September 11, 2007 48
Expert System Tools Are Not Design Rule Checkers! Design Rule Checkers Scan a board layout looking for design rule violations. Advantages Easier to understand what the software is doing Easier to use. Disadvantages Design rules don t apply in all situations Higher board cost to meet unnecessary design rules Will not detect problems that don t violate a pre-defined rule Santa Clara Valley Chapter Mtg. September 11, 2007 49
EMC Design Guideline Collection http://www.cvel.clemson.edu/emc/ Santa Clara Valley Chapter Mtg. September 11, 2007 50
Conclusion Four measurement and analysis technologies that were relatively unknown just a few years ago are becoming essential tools for EMC analysis. Near-Field Scanners Time/Frequency Domain Analyzers Mini-TEM Cells Expert System Techniques Santa Clara Valley Chapter Mtg. September 11, 2007 51