Debugging EMI Using a Digital Oscilloscope. Dave Rishavy Product Manager - Oscilloscopes

Transcription

1 Debugging EMI Using a Digital Oscilloscope Dave Rishavy Product Manager - Oscilloscopes 06/2009 Nov 2010 Fundamentals Scope Seminar of DSOs Signal Fidelity 1 1 1

2 Debugging EMI Using a Digital Oscilloscope l l l l l Background radiated emissions Basics of near field probing Frequency domain analysis using an oscilloscope l FFT computation l Dynamic range and sensitivity l Time gating l Frequency domain triggering EMI debugging process Measurement example 2

3 Background radiated emissions

4 Basic Principles: Radiated Emissions The following conditions must exist ı An Interference source A sufficiently high enough disturbance level in a frequency range that is relevant for RF emissions ı A Coupling mechanism transmits disturbance signals from the interference source to the emitting element ı An Emitting element (Antenna) capable of radiating the energy produced by the source into the far field 4

5 Basic Principles: Radiated Emissions The following conditions must exist ı An Interference source A sufficiently high enough disturbance level in a frequency range that is relevant for RF emissions ı A Coupling mechanism transmits disturbance signals from the interference source to the emitting element ı An Emitting element (Antenna) capable of radiating the energy produced by the source into the far field 5

6 Interference sources ı Fast switching signals within digital circuits Single-ended (asymmetrical) data signals Switched mode power supplies - harmonics Differential data signals with significant common mode component ı High-order harmonics decrease at 20 to 40 db/decade ı Structures on the PC board can begin to resonate at harmonic frequency 6

7 Inference Sources: Differential Mode RF Emissions ı Emission results when signal and return are not routed together ı Near field probe can detect this by positioning within the loop position of probe is critical 7

8 General steps to help reduce Differential Mode RF emissions ı Reduction of the loop area (i.e. closer routing of the forward and return conductors) ı Reduction of the current in the conductor loop (if possible without impacting the circuit operation.) ı Reduce the rise/fall times for the transmitted data signals ı Use filtering to eliminate higher-frequency signal components (limit the disturbance spectrum.) 8

9 Inference Sources: Common Mode RF Emissions ı Common problem in multilayer PC boards ı Caused by parasitic inductance in return path or asymmetrical transmission ı External cable acts as an antenna ı Rule of thumb for line length as an antenna: λ/10 not critical λ/6 critical 9

10 Common Mode RF Emissions Best Possible Differential mode transmission Undesired parasitic capacitance in return path Unbalanced parasitic terminating impedances 10

11 General steps to help reduce common-mode RF emissions ı Reduce the RFI current I CM by optimizing the layout, reducing the ground plane impedances or rearranging components ı Reduce higher-frequency signal components through filtering or by reducing the rise and fall times of digital signals ı Use shielding (lines, enclosures, etc.) ı Optimize the signal integrity to reduce unwanted overshoots (ringing) 11

12 Basic Principles: Radiated Emissions The following conditions must exist ı An Interference source A sufficiently high enough disturbance level in a frequency range that is relevant for RF emissions ı A Coupling mechanism transmits disturbance signals from the interference source to the emitting element ı An Emitting element (Antenna) capable of radiating the energy produced by the source into the far field 12

13 Coupling Mechanisms ı Three coupling paths: Direct RF emissions from the source, e.g. from a trace or an individual component RF emissions via connected power supply, data or signal lines Conducted emission via connected power supply, data or signal lines ı Coupling Mechanisms Coupling via a common impedance Electric field coupling parasitic capacitance between source and antenna Magnetic field coupling parasitic inductance between source and antenna Electromagnetic coupling far field coupling (greater than 1 wavelength) 13

14 Basic Principles: Radiated Emissions The following conditions must exist ı An Interference source A sufficiently high enough disturbance level in a frequency range that is relevant for RF emissions ı A Coupling mechanism transmits disturbance signals from the interference source to the emitting element ı An Emitting element (Antenna) capable of radiating the energy produced by the source into the far field 14

15 Emitting Elements (Antennas) ı Unintentional antennas in electronic equipment Connected lines (power supply, data/signal/control lines) Printed circuit board tracks and planes Internal cables between system components Components and heat sinks Slots and openings in enclosures ı Main factor is the length of the antenna with respect to the wavelength of the interference. Rule of thumb antennas with length less than λ/10 are not critical 15

16 Basics of near field probing

17 Near Field Definition Distance from DUT r Wave impedance E field H field Near field Transition Far field r = 1.6m for f > 30 MHz ı Sources with Low Voltage, but high current predominantly generate magnetic fields (e.g. terminated high speed signals) ı Sources with High Voltage, but low current predominantly generate electrical fields (e.g. unterminated signals) 17

18 EMI Debugging Example Oscilloscope and accessories R&S RTO R&S RTE Current clamp R&S EZ-17 Nearfield probes R&S HZ-15 E- and H-field 30 MHz 1 GHz Applicable from 100 khz Optional: R&S HZ-16 preamplifier 18

19 Magnetic and Electrical Near-Field Probes ı Basically the probes are antennas that pickup the magnetic & electric field variation ı The output Depends on the position & orientation of the probe 19

20 H-Field Probe H field Vo Current flow ı Maximum response with probe parallel with current and closest to the current carrying conductor ı Traces with relatively high current, terminated wires and cables 20

21 E-Field Probe Vo E field Current flow ı Maximum response with probe perpendicular with current and closest to the current carrying conductor ı Traces with relatively high voltage: unterminated Cables, PCB traces to high impedance logic (tri-state outputs of logic IC s) 21

22 Frequency domain analysis using an oscilloscope

23 Using an Oscilloscope for EMI Debugging ı Benefits Wide instantaneous frequency coverage Overlapping FFT computation with color grading Gates FFT analysis for correlated time-frequency analysis Frequency masks for triggering on intermittent events Deep memory for capture of long signal sequences ı Limitations Dynamic range No preselection No standard-compliant detectors (i.e CISPR) 23

24 Important Scope-Parameters for EMI Debugging Parameter Record length Sample rate Coupling Vertical sensitivity Color table & persistence FFT Span / RBW Signal zoom & FFT gating Description Ensure that you capture enough >2x max frequency, start with 2.5 GS/s for 0 1 GHz frequency range 50 Ω for near-field probes (important for bandwidth) 1 5 mv/div is usually a good setting across full BW Easily detect and distinguish CW signals and burst Easy to use familiar interface, Lively Update Easily isolate spurious spectral components in time domain 24

25 Frequency Domain Analysis FFT Basics FFT t s f FFT Integration time t int N FFT samples input for FFT Total bandwidth f s N FFT filter output of FFT ı N FFT ı f FFT ı t int ı f s Number of consecutive samples (acquired in time domain), power of 2 (e.g. 1024) Frequency resolution (RBW) = integration time FFT t sample rate f 1 = int f N s FFT 25

26 FFT as Basis for EMI Debugging with Oscilloscopes Conventional FFT Implementation on a Scope Time Domain t = 1/F s F max = F s /2 Frequency Domain x(t) S(f) S(f) t Data acquisition Windowing FFT f 1 f 2 f Zoom (f 1 f 2 ) f 1 f 2 Display f Record length T f = 1/T 26

27 FFT on the Rohde and Schwarz RTE and RTO Spectrum Analyzer Use Model ı Use model: Frequency domain controls time domain Time domain parameters automatically changed as Time Domain x(t) Zoom happens here before the FFT 500 MHz center, 10 MHz span: Fs = 1 GS/s vs 20 MS/s F s =2Β Data acquisition t HW Zoom (DDC) NCO LP necessary ı Downconversion FFT (DDC) zooms into frequency range before FFT Largely reduced record length, much faster FFT Decimation Windowing FFT S(f) Β=f 2 -f 1 f 1 f 2 Display f Frequency Domain Record length T f = 1/T 27

28 Measurement Consideration Gated FFT in the RTE and RTO Practical Time-Frequency Analysis Gated FFT: 50% overlap (default setting) One complete Time-Domain capture The Key to unraveling the time domain 28

29 FFT Gating 29

30 Measurement Consideration: Sensitivity Ability to detect weak Signals EMI tends to be weak and near field probes have low gain, the oscilloscope needs to be able to detect small signals over its full bandwidth 1mV/div Low Noise and High Sensitivity at Full Bandwidth 30

31 Signal to Noise and ENOB Higher ENOB => lower quantization error and higher SNR => Better accuracy l Thermal noise is proportion to BW. l An FFT bin is captures a narrow BW proportional to 1/ N FFT l Noise is reduced in each bin by a factor of 10 log10 l The limit approaches sum of all non-random errors. (Measurement induced errors are still present) 1 N FFT f FFT 31

32 Measurement Consideration: Signal to Noise >80 db 32

33 Measurement Consideration: Limit Lines ı Mask Tool Upper for limit line usage Mask definition in units of FFT Upper region mask acting as limit line Stop-on mask violation setting is very useful! ı 6 db EMI filter? Not critical for precompliance, will change results only slightly. 33

34 Measurement Consideration: Frequency Mask Triggering 34

35 EMI debugging process

36 The Problem: isolating sources of EMI ı EMI compliance is tested in the RF far field Compliance is based on specific allowable power levels as a function of frequency using a specific antenna, resolution bandwidth and distance from the DUT No localization of specific emitters within the DUT ı What happens when compliance fails? Need to locate where the offending emitter is within the DUT Local probing in the near field (close to the DUT) can help physically locate the problem Remediate using shielding or by reducing the EM radiation ı How do we find the source? Frequency domain measurement Time/frequency domain measurement Localizing in space 36

37 EMI Debugging Procedure Analysis steps EMI compliant testing / Test lab EMI Debugging / R&D A) Far-field measurement C) Reference measurement without DUT B) Know your DUT : List of potential interferer sources Source Clock frequency Ethernet PHY Voltage converter / power adapter Frequency e.g. 25 MHz + Multiples e.g. 125 MHz + Multiples broadband D) Interferer current measurement to find out the coupling type E) Nearfield probe to localize the interferer source F) Analysis of counter-measures 37

38 Observe the Spectrum While Scanning With a NearField Probe I) General Approach ı Wide Span scan fundamental of interfering signals are usually lower than 1GHz, a span of <1GHz is sufficient as a start 38

39 Observe the Spectrum While Scanning With a NearField Probe I) General Approach ı Wide Span scan fundamental of interfering signals are usually lower than 1GHz, a span of <1GHz is sufficient as a start ı Identify abnormal spurious or behavior and its location while moving the probe around 39

40 Observe the Spectrum While Scanning With a NearField Probe I) General Approach ı Wide Span scan fundamental of interfering signals are usually lower than 1GHz, a span of <1GHz is sufficient as a start ı Identify abnormal spike or behavior and its location while moving the probe around ı Narrow down to smaller span and RBW, change to smaller probe for better analysis 40

41 Measurement Example

42 Measurement Example IP Phone 42

43 Example: IP-phone Situation ı IP-phone components Complex processor unit DDR2 memory Ethernet Layer 2 Switch 2 x Gigabit Ethernet PHYs Several DC/DC Converters SPI-Interface to keyboard module Analog circuits (loudspeaker, Microphone) ı Failed in EMI compliant test Frequency (MHz) Level (dbµv/m) Limit (dbµv/m) Margin (db) Height (cm) Azimuth (deg) HOR HOR HOR Polarization 43

44 Far Field Test Result 44

45 RFI Current Measurement 375 MHz Spur Peak detect separates intermittent interference 45

46 Example: IP-phone Current-probing on interface lines 250 MHz 375 MHz 200 MHz 425 MHz ı Additionally detected emissions on following frequencies: CW: 375 MHz Broadband: 250 MHz 46

47 Example: IP-phone Nearfield-probing for source localization Lokalisierung DC/DC-converter no. 2 No significant emission Nearfield spectrum in the area of the processor module; among others 375 MHz interferer 47

48 Correlating Time and Frequency Domains 48

49 Example: IP-phone Results ı Interferer signal detected on interface lines The interferer is probably transferred via common-mode coupling ı Interferer sources localized DC/DC converter no. 1 Processor module respectively LAN PHY interfaces ı Analysis of layout and implementation of counter measures. 49

50 Debugging EMI Using a Digital Oscilloscope Summary ı If we can measure something in the far field, it must have an electric and magnetic near field source. ı The conditions required for a radiated emission allow us insight to track down a source and mitigate potential interferers. ı EMI Debugging with an Oscilloscope enables correlation of interfering signals with time domain while maintaining very fast and lively update rate. ı The combination of synchronized time and frequency domain analysis with advanced triggers allows engineers to gain insight on EMI problems to isolate and converge the source and solution quickly. ı Please see this shortcut to our application note for additional information: 50