EMC and Printed Circuit Board Design Design guidelines and concepts made simpler Part I Basics, Noise Sources, Tests

Transcription

1 Abstract The electronic products being developed now must pass stringent certifications before it hits out into the market. With the technology zooming further and the popularity of these reaching the conman with a much faster pace, the pressure to stay alive in the market has become high. With the full-fledged launch of wireless and other newer technologies, the EMC (Electro Magnetic Compatibility) problems become alarmingly high. In order to meet the stringent compliance requirements posed by FCC, IEC, CE, UL the design must be really robust. The objective is to make the user understand the basics of EMC and to apply these concepts while designing a successful printed circuit layout board. In this paper (Part I Basics, Noise sources, Tests) we will be going through the different standards of IEC (International Electrotechnical Commission) and their interpretation. The next series (Part II Design guidelines for EMI/EMS complaint PCB design) would deal with practical tips to designers to enhance their knowledge and help them design an EMI complaint PCB. EMC IEC Standards and Tests IEC is the governing body for the international EMC standards and conformity assessment for government, business and society for all electrical, electronic and related technologies. The definition of electromagnetic compatibility, as a one liner The system (device) should neither succumb to the noise generated by other devices / noise present in air (space) nor must conquer the other devices. Before we understand EMC and its effect better, lets understand the term noise that is being very commonly used. Noise is also an electromagnetic signal. Any signal that does not obey the specification and tries to hamper self and neighboring devices can be called a noise. But the definition is not yet complete. An honest specification-obeying signal can also be a noise. How? If a neighboring device develops a penchant towards this honest signal, they start talking (technically speaking cross talk / mutual inductance / electromagnetic radiation) among themselves and both the devices get affected. Girish Ramakrishnan Page 1

2 Electromagnetic Compatibility (EMC) is the ability of a system to perfectly perform the necessary functions as per the specification in the presence of noise a.k.a. EMS (Electro Magnetic Susceptibility). Noise can be internal or external and radiated or conducted. External noise is generally generated by other neighboring systems. Also the system must not influence noise (or generate noise) that could affect the other neighboring systems a.k.a. EMI (Electro Magnetic Interference). In order to determine the effective operation of the equipment in the presence of noise, EMC tests need to be conducted on the EUT (Equipment Under Test). These tests would determine how much noise the EUT radiates (emits) to the external environment and how immune is the EUT to the external noise. The EMC tests can be broadly classified as follows: Figure 1EMC Tests - A 35,000 feet view (not a complete list) If one observes carefully the figure 1, there are 2 major building blocks viz. the radiation and conduction. The task for the systems engineer would be to understand which factor amongst this two is hampering his product. If, he were able to narrow down on these, he would at least be in a position to understand the problem definition better. Radiation is the propagation of electromagnetic waves through air. Conduction is the propagation of electromagnetic waves through a physical medium. Girish Ramakrishnan Page 2

3 In the interference tests (radiated and conducted) the level or the intensity of the electromagnetic waves generated by the EUT and their propagation into different mediums is studied. In the susceptibility tests (radiated and conducted) the effect of the electromagnetic waves received through different mediums and the sensitivity, performance of the EUT is verified. Due to the effect of noise, the system can become erratic, might often get reset, may not perform the desired operation, and in the worst case the system will stop functioning. Based on this, IEC has classified the performance criteria into 4 levels. Level A, Level B, Level C and Level D. Level A Equipment Under Test (EUT) performs as per specifications without any deviations or errors during and after the test. Level B EUT malfunctions temporarily during test but after the test recovers automatically to normalcy without user intervention. Level C- EUT malfunctions temporarily and does not recover automatically during or after the test and requires user intervention to bring it back to normalcy. Level D EUT malfunctions or degrades catastrophically which is not recoverable damaging the software, hardware or data and cannot return to its normal working condition. Complete failure. IEC does not recommend any specific pass condition. A pass condition for a particular EUT is generally derived by an agreement between the test engineering team and the design team. Obviously, while the agreement is being drafted, the parameters, which would be considered, include the compliance requirement that the product needs to undergo and the application of the product. Severity levels of compliance might vary depending on the environment, regulatory requirements, etc. Sources of EMC AC/DC power lines Motors Radios / Television transmitters Welding machines, Electrical Drills Girish Ramakrishnan Page 3

4 Communication networks (GSM, CDMA) Lighting Switches operating inductive load Electrostatic Discharge The list is never ending. These listed above are very common noise sources. They can cause EMI and EMS on your device. EMI Tests The radiated test governed by IEC CISPR 22 standards, measures the amount of noise that the EUT emits to the air that can be picked up by the other neighboring devices. The conducted test governed by IEC standards, measures the amount of noise that the EUT emits to the main power (AC or DC power supply). If the amount of conducted noise emitted is much more than the desired level it would affect the other neighboring systems sharing the same power supply. The objective of these tests is to determine the level of noise at different frequencies and to verify if they satisfy the criterion drawn by the IEC on the noise levels. The noise limits are classified into two levels, viz. Class B and Class A. Class B level is defined for the equipments that fall into the domestic category and Class A level is defined for industrial equipments. The Class A level is much more stringent than the Class B level. The red line in the below figures shows the maximum acceptable noise level (in dbµv/m) recommended by IEC and the waveforms below the red line (in the range of 15dBµV/m to 20dBµV/m) indicates that the EUT is under safe working condition and has passed the radiation tests. Girish Ramakrishnan Page 4

5 Figure 2 Typical Radiation Emission Waveform Frequency Range: MHz Figure 3 Typical Radiation Emission Waveform Frequency Range: 150KHz - 30 MHz There are many tutorials available in the World Wide Web which speaks a lot about the test equipments viz., LISN (Line Impedance Stabilization Network), spectrum analyzer, EMI receiver, test procedures, types of antennas, anechoic chambers, GTEM (Gigahertz Transverse Electro Magnetic) cell, OATS (Open Area Test Site), etc and hence I am not going into the details of these. TIP: A general guidance before starting any EMI test with the EUT its better to measure the ambient noise present at that particular moment so that after the test if required the ambient noise can be subtracted from the actual noise measured during test to give accurate results. Girish Ramakrishnan Page 5

6 EMS Tests EMS tests analyze the performance of the EUT in the presence of external noise sources. They can also be of 2 types viz., radiated component and conducted component. The EUT is subjected to the external radiated noise and the performance is analyzed. This test is called radiated, radio frequency, electromagnetic field immunity test and governed by IEC standard. This test is to verify the susceptibility levels of the EUT under the influence of electromagnetic radiation / fields generated by radio/television transmitter, GSM/CDMA mobile communication networks and other external sources that emit electromagnetic waves. If the EUT operates in the same range as that of a FM transmitter or if the transmitter operates in the nth harmonic of the EUT frequency, interference is most likely to happen. Depending on the type of the EUT IEC has set 4 different test levels, viz. Level 1, level 2, Level 3 and Custom level. These levels mention different field strength (measured in V/m) of the EM waves that are generated by the test equipment. The unmodulated signal with the IEC specified field strength is 80% amplitude modulated with a 1KHz sine wave to simulate the actual EM radiation. The conducted section of the electromagnetic susceptibility tests consists of many tests. This paper would deals with only a few tests and not the complete list though as listed by the IEC. Electrical Fast Transients, IEC High Energy Surge, IEC Electrostatic Discharge, IEC Electrical Fast Transients (EFT) In this test, a number of bursts at a specified voltage level is coupled with the power supply and control lines. The standard format of EFT is bursts coupling with the power supply (line, neutral and earth for a single phase system). Another flavor of EFT is defined by the IEC where the I/O signals can be coupled with the noise. This is called EFT I/O. Both the mentioned tests are defined in IEC standards. Spurious radiations are caused by devices such as welders, thyristors, fluorescent lights, switches operating inductive loads. This interference manifests itself as conducted electrical Girish Ramakrishnan Page 6

7 interference to be more specific the EFT. Lets say a welding machine and the EUT are connected to the same power supply. The noise that is put back by the welding machine into the power supply might affect the EUT. The IEC standard noise generator simulates the noise that is put back by the welding machine whose waveform is depicted in figure 4. Number of the bursts would be generated while running a 1-minute test of EFT. The IEC explains the test procedure. The noise signal of different magnitude from 500V until 4000V in both the polarities can be typically coupled with the power lines, as explained earlier. The rise time of each pulse in the burst is 5 ns and hence the name fast transients. If the noise pulse synchronizes with the rise time of a signal in the printed circuit board and if they are not EFT immune, it can create havoc. The common problems faced during an EFT test are: frequent resets, EUT might stop functioning, weird behavior, not be able to turn on the loads, the HMI (Human Machine Interface) or the UI (User Interface) corruption. To successfully pass the EFT power line test, then the power supply must have all necessary protection circuitry and must be made robust. The common protection devices for EFT could be filters, transient voltage suppressors, ferrite beads, etc. A perfectly designed RC filter can be a low cost and highly robust solution for EFT. Study of the waveforms shown in figure 4 suggests each burst consists of many EFT pulses. The duration of each burst is 15ms and the period of each burst is 300ms. Voltage magnitude less than 2000V are tested for a repetition frequency of each pulse at 5KHz and more than 2000V at 2.5KHz *. * Note: As per IEC release A much deeper look at the waveforms, suggests the approximate number of pulses at a particular instant t. For a repetition frequency of 2.5KHz there would be approximately 8000 EFT pulses in each EFT burst (each EFT pulse is represented in the zoomed view, figure 4 and called as standard EFT waveform). For burst duration of 15ms there would be 38 EFT bursts, which means for burst duration of 15ms there could be approximately EFT pulses. This explanation is provided to the user to imagine the intensity of this test. Girish Ramakrishnan Page 7

8 Figure 4 EFT Waveforms Magnified View The same set of test waveforms is also applicable for EFT data I/O coupling. The EFT noise is capacitive coupled with the power supply. A 33nF capacitor is used for coupling with AC or DC mains. Generally these capacitors are included in the IEC certified standard noise generators. In case of data line I/O coupling a separate capacitive coupling clamp is used. The I/O signals are allowed to pass through the capacitive clamp and the effect of the EFT noise on the EUT is studied. The capacitive coupling clamp is about one meter long. In many of the occasions an I/O signal harness may not run 1 meter. In that case one may not be able to place the I/O harness in the capacitive coupling clamp and perform the EFT I/O test. IEC recommends an alternate strategy to test in such cases. The harness can be wrapped in a tape or a conductive foil. The capacitance of this coupling arrangement should be equivalent to that of a standard coupling clamp. To understand this better refer figure 5. The author Girish Ramakrishnan Page 8

9 tried such coupling ideas in practical applications and was able to observe nearly realistic effects, however lab made coupling ideas are just a work around solution and proper care should be taken in experimenting such methods. Figure 5 Capacitive Coupling on I/O lines - alternate arrangement High Energy Surge Surge is a transient wave characterized by rapid changes in the current or voltage flow along a PCB trace or wire. Switching loads, switching power grids or lightning generates high-energy surges. They are typically slow rising signals and last a few milliseconds. The difference between the EFT and the surge is, EFT is a fast rising, low energy signal while the surge is a slow and high energy one. The surge test is governed by IEC IEC defines the test as application of the high-energy surge to the power mains or I/O lines. The IEC has defined different threat levels and the surge tests are done as prescribed by the IEC standards. Surges are of different types: Combination Wave 1.2/50µs (a.k.a. open circuit voltage test), 8/20µs (a.k.a. short circuit current test) defined by the IEC In this test the surge can be applied in the power lines or I/O signals at different firing angles (phase shifts), in both the polarities and common / differential modes. This test is to simulate the condition of surge due to lightning, switching loads, short circuits and arching. The standard waveform of the surge noise signal generated for this combination wave test is shown in figure 6. Ring Wave 0.5/10µs 100KHz defined by the IEC This test is to simulate the condition of surge due to ringing transients in over-head transmission lines. 10/700 Telecom Surge defined by the IEC This test is to simulate the lighting transients on telephone lines. Girish Ramakrishnan Page 9

10 1MHz damped sinusoidal defined by the IEC This test is to simulate the oscillatory transients due to switching of power grid, inductive or capacitive loads. Based on the EUT s application, the type of the surge to be applied for the test needs to be decided. Home appliances can undergo combination wave and ring wave tests. Telecom equipments can undergo telecom surge, home appliances which have reactive loads (inductive, capacitive) whose power supply ruggedness needs to be verified can undergo combination, ring wave tests along with damped sinusoidal tests. The test procedure for each of the tests is well defined and explained in their respective IEC standards. Figure 6 Surge Waveform Capacitive coupling via 9µF between line and earth or 18µF between line and line, capacitors are required for coupling surges to AC or DC power mains. These coupling capacitors are typically included as a part of the IEC certified surge simulator (noise generator). Capacitive Coupling via 0.5µF capacitors, as well as via alternative devices such as surge arrestors, are used to couple surges onto data, I/O and telecommunication lines. Surge arrestor is the preferred coupler device due to the undesirable loading effects of capacitors particularly at frequencies greater than a few kilohertz. They are typically included as part of the IEC certified surge simulator. A common protection device for Surge could be MOV (Metal oxide varistor), thyristors and voltage suppressors. Girish Ramakrishnan Page 10

11 Electrostatic c Discharge Electrostatic discharge is the transfer of static charge between two objects. Such a sudden discharge of charges can sometimes be catastrophic. The scope as defined by IEC is the static charges generated by humans during an interaction with electronic equipment. The physics of ESD defines that in humid conditions the effect of ESD would be much lesser than the effect in dry conditions. The effect of ESD on the electronic equipments used in the regions near the equator of the earth generally is less. But this is at a system level. If we look ESD into a microscope level, viz. for semiconductors the perception will not be the same. IEC defines the standard testing from the macroscopic level (system) in the standards. IEC recommends two methods for discharging the ESD. One is the contact discharge while the other is the air discharge. The test waveform depicted in figure 7, is common for both air and contact discharge. The detailed test procedure (how to hold the ESD gun, typical distance between the tip of the gun and the EUT, etc) is explained in the IEC standards. ESD is being tested for a variety of electronic products all over the globe and hence IEC has defined different threat levels categorized in accordance with the industry and type of device. There are two different coupling methods viz. direct and indirect. Direct Coupling will discharge directly to the EUT, either in an Air Discharge mode (necessary for products having few or no metal surfaces) or Contact mode (the preferred method). Indirect Coupling simulates an ESD event that causes a radiated field to be emitted in the vicinity of the victim equipment. This is accomplished using Vertical Coupling Planes (VCP) and Horizontal Coupling Planes (HCP). Refer to figure 8 for a better understanding of vertical coupling planes and horizontal coupling planes. Figure 7 ESD Waveform Girish Ramakrishnan Page 11

12 Figure 8 depicts ESD noise coupling on VCP, HCP and nothing more. Figure 8 HCP and VCP References International Electrotechnical Commission - IEC Series of Standards: Electromagnetic Compatibility "Requirements for Household Appliances, Electric Tools and Similar Apparatus". About the Author Girish Ramakrishnan, Sr. Design Analyst, Satyam Computer Services Ltd., Hyderabad, India, holds a Masters Degree in Microsystems Engineering from University of Applied Sciences Furtwangen, Germany. He has over 6 years of experience and worked with Atmel Germany, LG Electronics India prior to joining Satyam. He has vast experience in PCB Designs, EMI EMC design practices and EMI EMC testing. He is a certified Six Sigma Green Belt. Girish Ramakrishnan Page 12