Understanding and Optimizing Electromagnetic Compatibility in Switchmode Power Supplies

Transcription

1 Understanding and Optimizing Electromagnetic Compatibility in Switchmode Power Supplies 1

2 Definitions EMI = Electro Magnetic Interference EMC = Electro Magnetic Compatibility (No EMI) Three Components of EMI: Noise generator (source) Noise transmission (coupling) Noise receptor (victim) 2

3 Understanding EMI Electromagnetic Fields Caused by Changing Currents and Voltages Maxwell s Equations: di e = M, i = dt C dv dt Slower Rise/Fall Times (with tradeoffs) Low Energy Phenomenon At 1 Mhz, 20 nw can fail FCC limits 3

4 Categorizing EMI EMI Conducted (<30 MHz) Radiated (>30 MHz) Differential Common Mode Mode Magnetic Electric System Usage: Class A - Industrial and Commercial Class B - Includes Residential 4

5 EMI Specifications USA - FCC: CFR Title 47, Part 15 (etc.) Europe - IEC: EN50081 (and others) French - CISPR: Publication 22 Frequency Ranges (FCC) - Conducted EMI: Radiated EMI: 450 khz to 30 MHz 30 MHz to 1 GHz Note: Measurement dependencies on operating conditions 5

6 Conducted Noise Limits (FCC vs CISPR) FCC CLASS A Voltage - dbuv CISPR CISPR CLASS B 45 FCC Frequency - MHz 6

7 Radiated Noise Limits (FCC vs CISPR) Measuring Distance = 10 m FCC Field Strength - db uv/m CLASS A CLASS B CISPR FCC CISPR Frequency - MHz 7

8 Measuring Conducted Noise Line Impedance Stabilization Network (LISN) Power Source 50 µh Power Supply Input Ground 10 µf 50 Ω 10 nf to 330 nf* To Spectrum Analyzer * Capacitor value determined by lowest specified frequency 8

9 Elements of Conducted Noise LINE Two LISN Circuits CM GROUND DM SMPS OUTPUTS NEUTRAL CM Spectrum Analyzer 50 Ω CM CM and DM add vectorially EMI (line) = CM + DM EMI (neutral) = CM - DM 9

10 Sources of Differential Mode Noise INPUT NOISE OUTPUT NOISE Switching action causes current pulses at input and output. 10

11 Sources of Common Mode Noise Capacitive Currents Primary to Secondary Input Noise Output Noise Chassis Ground Chassis Ground Chassis Ground Capacitive Currents Direct to Chassis 11

12 Combating Differential Mode Noise POWER SOURCE Line Return DM Input Filter Positive Input Negative Input SMPS + VOUT - Ground 12

13 Basic Ideal Filter 20 µh V IN V OUT 3200 µf db 20 µh 3200 µf Attenuation - db 40 db 60 db 80 db (a) 100 db 120 db 100Hz 1 khz 10 khz 100 khz 1 MHz 10 MHz 100 MHz 1 GHz Frequency - Hz 13

14 Actual Filter with Parasitics 20 µh V IN V OUT 500 pf ESL 16 nh + ESR 0.02 Ω 3200 µf 0 16 nh 500 pf 20 db 20 µh 3200 µf 0.02 Ω 3200 µf (b) Attenuation - db 40 db 60 db 80 db (a) (b) 0.02 Ω 16 µh 500 pf 20 µh 100 db 120 db 100Hz 1 khz 10 khz 100 khz 1 MHz 10 MHz 100 MHz 1 GHz Frequency - Hz 14

15 Four Paralleled Capacitors Reduces ESL 20 µh V IN V OUT 500 pf 16 nh 16 nh 16 nh 16 nh 0.08 Ω 0.08 Ω 0.08 Ω 0.08 Ω µf µf µf µf 0 4 nh 500 pf 20 db 20 µh 3200 µf 0.02 Ω 3200 µf (b) (c) Attenuation - db 40 db 60 db 80 db (a) (c) 500 pf 20 µh 100 db 0.02 Ω 4 nh 120 db 100Hz 1 khz 10 khz 100 khz 1 MHz 10 MHz 100 MHz 1 GHz Frequency - Hz 15

16 Improved Inductor Reduces Shunt C 20 µh V IN V OUT 50 pf 16 nh 16 nh 16 nh 16 nh 0.08 Ω 0.08 Ω 0.08 Ω 0.08 Ω µf µf µf µf 0 4 nh 50 pf 20 db 20 µh 3200 µf 0.02 Ω 3200 µf (b) (c) Attenuation - db 40 db 60 db 80 db (a) (d) 100 db 0.02 Ω 4 nh 50 pf 20 µh 120 db 100Hz 1 khz 10 khz 100 khz 1 MHz 10 MHz 100 MHz 1 GHz Frequency - Hz 16

17 Adding a Second Stage Inductor 5 pf V IN 20 µh 1 µh V OUT 50 pf 16 nh 16 nh 16 nh 16 nh 0.08 Ω 0.08 Ω 0.08 Ω 0.08 Ω µf µf+ 800 µf µf 0 8 nh 5 pf Attenuation - db 20 db 40 db 60 db 80 db 20 µh 3200 µf 0.02 Ω 3200 µf (a) 0.08 Ω 1 µh (b) (e) (c) (b) (c) (d) 5 pf 1 µh (e) 8 nh 50 pf 100 db 0.02 Ω 4 nh 50 pf 20 µh 120 db 100Hz 1 khz 10 khz 100 khz 1 MHz 10 MHz 100 MHz 1 GHz Frequency - Hz 17

18 Reducing Inductor Parasitic Shunt Capacitance S Windings F Bobbin Core 18

19 Filter Resonance Natural Resonant Frequency = 2π 1 LC Three Potential Problems: 1. Step application of input voltage could ring to 2V P. 2. High frequency noise at input could be amplified by Q of filter. 3. Filter output impedance rises at fr with potential oscillation with Z IN of converter. 19

20 Filter Damping Undamped Filter Damping Components L LINE SOURCE C R D 2 C D C L C POWER SUPPLY RETURN 20

21 Combating Common Mode Noise +200 VDC V 400 V 5 µs I 100 ns 20 ns +48 ma 12 pf? 8 ma I RMS = 9.6 ma (Typical TO-220 insulator capacity is 12 pf, f s = 200 khz) 21

22 Common Mode Noise Analysis LISNs SMPS PWR 12 pf V N RET 60 V RMS 600 khz 25 Ω V N V N 50 Ω 50 Ω 400 V 200 khz 12 pf 3rd harmonic equivalent noise voltage circuit I CM V N = 68 mv FCC Limit (class A) = 1.0 mv Required noise filter attenuation: 37 db at 600 khz 22

23 Achieving 37 db Attenuation With Series Inductor Co < 0.17 pf! 12 pf 419 mh 0.95 mv V N 60 V RMS 600 khz (Xc = 22.1 kω) (X L = 1.58 MΩ) 2 X 250T) 25 Ω Allowable parasitic capacitance is unrealistic. 23

24 Achieving 37 db Attenuation With Shunt Capacitor 60 V RMS 600 khz 12 pf (X C = 22.1 kω) 760 nf (X C = 0.35 Ω) ESL < 93 nh 25 Ω 0.95 mv V N Required capacitor will not meet safety specifications. 24

25 Optimum Solution Uses Both L and C 153 mv 12 pf Co < 66 pf 1.07 mh 0.95 mv V N 60 V RMS 600 khz 4.7 nf (X C = 56 Ω) (X L = 4.03 kω) 2X13T) 25 Ω 25

26 Complete Input Filter for Both DM and CM Conducted Noise Differential Mode Inrush Limiter L N L c1 AC Line C c2 Cd3 C c1 C d2 Discharge C N C d1 DC Bus C c2 R N L c1 C c1 L d1 Final Cleanup Common Mode Differential Mode Notch filter for f S 26

27 Minimizing CM Noise Injection Heat Sink Tied to PCB Ground (return) Insulators Case at HF AC PCB Bracket Contacting PCB Ground PCB Mounted Heat Sink Chassis Grounded Heat Sink Shielding Heat Sink Bracket 27

28 Electrostatic Shielding CORE Use of Primary Shield ALTERNATE SHIELD CONNECTION V Unshielded Transformer Correct Incorrect! Also called a Faraday shield Connect to V+ if turn-off is fastest, to return with faster turn-on 28

29 Radiated EMI Noise is easily transformed back and forth between conducted and radiated form Conductors become antennas and antennas become receivers Testing more difficult Frequency > 30 MHz Test environment and fixturing is critical 29

30 Radiated Noise Measurements Power Source SMPS "Load" Test fixture on rotating turntable Antenna on variable height vertical support Antenna Spectrum Analyzer 30

31 Characteristics of Radiated Noise dv E = C dt Minimize high dv/dt on large surfaces di H = M dt Minimize high di/dt in conductive loops λ < 2 π Electric and magnetic fields act independently λ > 2 π Electric and magnetic fields merge Electric Field: Magnetic Field: Near Field: Far Field: At 1 MHz, λ = 300 meters 31

32 Potential Electric Field EMI Sources E E E LARGE AREA, HIGH dv/dt ANTENNAS Shielding possible 32

33 Magnetic Fields from Transformers and Inductors High di/dt Switching Loops I I H H Stray Transformer and Inductor Fields Shielding difficult 33

34 Potential Magnetic Field EMI Sources Rectifier Rectifier Transformer Transformer Large Loop Area With High di/dt Wide, Closely Spaced Copper Straps Worst case! Much Better! 34

35 Magnetic Fields from Leakage Inductance SEC PRI Core PRI SEC P SEC P Core P SEC P Leakage fields radiate 1 with intensity of 3 d Opposing fields tend to cancel 1 with intensity of 4 d 35

36 Minimizing Stray Magnetic Fields in Transformers Continuous copper strap around both windings and core. Converts stray magnetic fields to eddy current. Eddy current creates a canceling magnetic field. 36

37 Minimizing Stray Magnetic Fields in Inductors Core Winding Coil Core Poor Construction Techniques 37

38 Summary Presented a general overview Defined various categories of noise Described measurement techniques Discussed ways to minimize noise generation Did not cover shielding or techniques to minimize susceptibility Should mention frequency modulation Valuable additional references listed 38