Academic Year 2015-2016 ELEC 0017: ELECTROMAGNETIC COMPATIBILITY LABORATORY SESSIONS V. BEAUVOIS P. BEERTEN C. GEUZAINE 1
CONTENTS: EMC laboratory session 1: EMC tests of a commercial Christmas LED light set... p. 3 CEM laboratory session 2: Reverberation chamber and Mobile phone measurements... p. 6 EMC laboratory session 3: EMC constraints for integrated circuits and PCBs... p. 10 2
EMC laboratory session 1 EMC tests of a commercial Christmas LED light set 1. Visit of the ACE semi-anechoic chamber What is an anechoic chamber and why is the ACE chamber called semi-anechoic? What kind of material is ferrite? Why is it used in anechoic chambers? What are the blue cones on the walls of the chamber and what are they used for? Describe briefly how to select which standard(s) must be applied when first faced to the Equipment-under-Test (EUT). There are many different types of antennas, what are the most relevant parameters to consider when selecting an antenna for an EMI experiment? Who needs to have EMC tests done and why? 3
2. EMC testing of the Christmas light set a. Emission Which standard(s) did you choose for the emission tests? Why? What is a LISN? Why is it used for the EMI tests? Why are average and peak values both measured? Measurements: Does the light set pass all the emission tests? Justify. b. Immunity Which standard(s) did you choose for the immunity tests? Why? What are the 3 different performance criteria? What phenomena cause the existence of surges in the network? Injected currents: why are they considered as conducted-only below 80 MHz and radiated above that frequency? Measurements: Does the light set pass all the immunity tests? Justify. What happens when EUTs fail the EMC tests? If the Christmas light set failed one or several EMC tests, could something be done by the manufacturer to improve the light set performances? 4
3. Other examples: Could our light set be installed in a train? Is a low cost supermarket voltmeter submitted to the same EMC tests as a high precision laboratory spectrum (both are laboratory equipment)? Why? 4. Electric field measurement under high voltage cables What are the main challenges of on-site open-air measurements? What is the order of magnitude of the electric field you measured under the power line? Can you compare it to other relevant electric field values? What is the order of magnitude of the magnetic field under the power line? 5
CEM laboratory session 2: Reverberation chamber and Mobile phone measurements Follow the instructions below and briefly answer the following questions in italic (a couple of sentences maximum). 1. The reverberation chamber (RC) What is a reverberation chamber? Compare it to the anechoic chamber you visited last time. Why are the walls of the RC made of metal? What is the role of the stirrer? Why are two calibrations necessary before running tests in the RC? What is the particular interest of RCs for EMC tests? Explain the reasons why the RC must be operated within the [200 MHz 18 GHz] range. 2. Shielding effectiveness of common materials What does the term shielding effectiveness (SE) mean and how is it calculated? Draw a simple diagram of the experimental set-up used to measure the SE. 6
How did you choose the frequency? How did you select the antenna to use for emission and for reception? Why is the voltage on the ESU/spectrum scaled in dbµv? Check that the noise level is well below the signal level when the antenna is emitting and no obstacle is placed between the antennas. Make a table listing all the materials you placed between the antennas and the resulting voltage measured at the reception antenna. Calculate the SE for all cases. What material exhibited the larger shielding effectiveness? Why? 3. Mobile phone measurements: a. Power of a single time slot Connect the GSM antenna to the spectrum. Put the mobile phone in communications (call 04/366.36.78 e.g.). Press the PRESET key to set the spectrum to its default state. Find the frequency of the communication channel with the highest power: set the center frequency (FREQ key) to 900 MHz with a span (SPAN key) and observe the peaks, use the markers to find the exact frequency corresponding to the highest peak (f max ), put the trace on MAX HOLD (in the TRACE menu) for better visibility. What is the frequency corresponding to the maximum peak you can observe (f max )? Put the trace back to CLEAR/WRITE and change the center frequency to the value of f max with a span 0 Hz (or press the ZERO SPAN key) to use the spectrum as a single frequency scope. Set the reference level to 10 dbm (check if this value is appropriate) using the AMPT key. Set the sweep time to 1 ms using the SWEEP key (Sweeptime Manual). To isolate and stabilize a single burst, adjust the trigger parameters using the TRIG menu: Press the Trg/Gate Source key and use the arrows to select the Video mode. Then set the Trg/Gate Level to 70%. The trigger level is visible on the display as a red horizontal line labeled with the absolute value for the trigger threshold. Plot the shape of the GSM time slot you isolated on the spectrum. To use the spectrum to measure the power of the burst, open the Time Domain Power submenu in the MEAS menu. Switch the Limits to ON. Adjust the right and left limits (vertical lines) of the burst to the beginning and the end of the burst using the Right Limit and Left Limit softkeys and the rotary knob. The power of the burst value can be found on the spectrum display in dbm (RMS). What is the duration of the burst? Measure the corresponding power. 7
b. Rise and fall times of the GSM burst To visualize the rising edge of the burst in high time resolution, first turn of the previous measurement parameters by pressing the All Functions Off key in the MEAS menu. Set the sweep time to 100 µs and set the Trigger Offset (TRIG menu) to -50 µs. Use the markers to estimate the value of the rise time of the burst. To visualize the falling edge, set the Trigger Offset to 500 µs. Use the markers to estimate the value of the fall time of the burst. Measure the rise and fall times of the burst. c. Signal-to-Noise ratio of the GSM burst Reset the spectrum to its default state then set the center frequency back to f_max and return to zero span mode. Set the resolution bandwidth to 1 MHz using the Res BW Manual key in the BW menu. Set the reference level to 10 dbm (AMPT menu). Set the sweep time to 2 ms (Sweeptime Manual key in the SWEEP menu). Set the Trg/Gate Source to Video, the Trg/Gate level to 70% and the Trigger Offset to -1 ms (TRIG menu). The GSM burst should be well displayed in the right-hand half of the spectrum screen. Measure the power of the burst according to the previous instructions. Calculate the signal-tonoise ratio of the burst. To simply and efficiently measure the power of the noise during an equivalent time interval, simply set the Trg/Gate Polarity to Neg (the burst is now in the left-had half of the screen, outside the measurement zone previously defined). d. Radiation Diagram of the GSM and influence of EM shields/patches 8
Reset the spectrum to its default state then set the center frequency back to f_max and return to zero span mode. Set the reference level to 10 dbm, set the sweep time to 2 ms and set the Trg/Gate Source to Video, the Trg/Gate level to 70% and the Trigger Offset to -1 ms. Adjust the left and right limits and the power of the burst should be displayed on the screen (as in the second experiment). Rotate the GSM holder by increments of 60 degrees and note the corresponding power values. Repeat the previous point for each patch or shield and note the values. Plot in polar coordinates a simplified radiation diagram of the mobile phone radiation power in the time slot selected previously, do it again using the electromagnetic shield and with the patch. Explain the results in a few words. 9
EMC laboratory session 3: EMC constraints for integrated circuits and PCBs Read again the slides on Design Rules for electronic circuits and PCBs (part I and II) from the EMC course follow the instructions below and briefly answer the questions in italic. Experiment 1: PCB track loops and unshielded cables Requires: - PCB [5.1]: single loop track to a 50 ohm charge, - PCB [5.2]: loop track with a series capacitor, - PCB [5.3]: two straight parallel tracks, - PCB [5.4]: single straight track with a background plane, - a digital oscilloscope, - a signal generator. Steps: - Connect both ends of a BNC cable to the output of the signal generator. Configure it to generate a squarewave signal at 1 MHz. Observe the shape of the signal on the digital oscilloscope. Isolate a single square. - Use now a 50 ohms T-termination and describe the change in the observed signal. - Connect a banana cable to the signal generator. - Place the PCB [5.1] above the banana cable (using a polystyrene stand), observe the signal on the digital oscilloscope and memorize the trace. Try to explain the deviations from a perfect square form. - Repeat the previous step for cards [5.2], [5.3] and [5.4]. - Compare the four traces on the digital oscilloscope. - Why is a banana cable used here, instead of a coaxial one? 10
Experiment 2: Decoupling capacitors and resonance frequency Requires: - PCB [12]: two loop tracks with series capacitors and one switch button used to select either the small or the large loop, - PCB [5.1]: single loop track to a 50 ohm charge, - a digital oscilloscope, - a signal generator. Steps: - Connect the BNC connector of PCB [12] to the output of the signal generator with a square wave of 2 MHz (15V). - Using PCB [5.1] connected to the digital oscilloscope as a crude magnetic field sensor (justify?), observe and explain the change of the signal behavior when the switch is on 0 (large loop) or on 1 (small loop). - Considering that both capacitors have a 100nF capacitance and that a single conductor has an inductance of 1µH/m 1, can you estimate the resonance frequency of both circuits? Experiment 3: Coupling by common impedance Requires: - PCB [X(top)], - a digital oscilloscope, - a DC voltage source. Steps: - Two components: a N555 and a 7400 must be connected to the ground, using the first switch you can connect them using independent ground lines or a common ground line. The common ground line can be selected using the second switch to be either short and large or long and thin. Observe the signal from the V out connector on the oscilloscope in the three configurations. - Compare the three traces and explain. 1 Source : «Electro magnetic compatibility and printed circuit board (PCB) constraints, Philipps Semiconductors, application note, June 1989. 11
Experiment 4: Influence of slots in the ground plane Requires: - PCB [13/7]: single track with complete ground plane, - PCB [13/5]: single track, ground plane with a slot, - PCB [13/6]: single track going around the ground plane slot, - a spectrum analyzer with tracking generator, - a magnetic field probe. Steps: - Connect the BNC connector of PCB [13/7] to the tracking generator output. Connect the magnetic field probe (using the preamp module) to the input of the spectrum analyzer. Place the probe a few centimeters above the PCB and record the trace to memory. - Repeat the previous step for PCB [13/5] and [13/6], place the probe above the slot and above intact ground plane for comparison. - Compare the results and explain. Experiment 5: Impedance matching of transmission lines Requires: - PCB [4], 8 BNC connectors for 8 different terminations: four resistors (10, 50, 100 and 680 kohms), one open, one Zener diode and two capacitors (100nF and 1nF). - a vector network analyzer. Steps: - Calibrate the VNA using the open, short and match standards(full frequency sweep), - Connect the open termination of the PCB [4] to the VNA and display the reflection parameter S11 using the Smith chart option. Explain the differences between the measured signal and a theoretically perfect open. Copy the trace to memory. - Repeat the previous step for both capacitors and compare those traces with the open termination case. Compare them also using the magnitude and phase formats. - Delete the previous traces except the one corresponding to the open termination. Connect the 680-ohm resistor and compare the results with the open. - Connect the 50-ohm resistor, observe the signal on the Smith chart but also in the SWR (standing wave ration) format, explain. Copy the trace to memory. - Connect the other resistor terminations and compare their traces in the SWR display. - Connect the capacitor terminations and compare the SWR with the resistors results. 12
Experiment 6: Influence of the choice of the logic family of the components: Requires: - PCB [1]: four identical circuits using different logic family for the 7400 2 Steps: component: (7400, 74LS00, 74HC00 and 74HCT00) with filtered and non-filtered output for each component. - an oscilloscope, - a spectrum analyzer, - a signal generator. - Connect clock input of PCB [12] to the output of the signal generator with a square wave of 1 MHz (5V). Verify the signal shape using the oscilloscope, measure the rise and fall times. - Connect the first output of the PCB to the spectrum analyzer (150 khz-80 MHz sweep and a reference level of 87 dbµv). Adjust the spectrum parameters to obtain an optimum signal. Measure the value of the first peak (1MHz). - Repeat the previous step for each BNC connector. For each logic family, compare the filtered and non-filtered output signal and the values of the 1MHz peak. - Compare the output signals of the four circuits and draw conclusions about the logic family used. - Using the oscilloscope, measure for each output the rise and fall times of the output square wave. What can you conclude from those measurements, about the different technology used? 2 They include four two-input NAND gates using TTL or different technologies but in a TTL-compatible format; 7400 is standard TTL technology, 74LS00 stands for Low-power Schottky, implemented using the same technology as 74S but with reduced power consumption and switching speed, 74HC00: High-speed CMOS, similar performance to LS and 74HCT00: High speed CMOS, compatible logic levels to bipolar parts. (source: http://en.wikipedia.org/wiki/7400_series). 13