Automotive Capacitive Sensing Electronic Compliance
Hurdles continue to emerge for those designing automotive electronic subsystems, with complexity growing as radiation sources proliferate.
Product compliance in electronic systems is governed by electromagnetic compatibility (EMC) tests. Qualifying a specific set of tests is necessary for the sale and use of an electronic product in a particular geographical region. For automotive electronic systems, special EMC tests are defined, since the electronic subsystem needs to survive and operate flawlessly in the vicinity of other noisy electrical devices. The vibrations and temperature ranges that exist in an automotive environment also require an additional degree of qualification, defined by a separate IC-level qualification process such as AEC-Q100.
“This also brings out the need to automate the process of EMC testing since a typical EMC test now takes several hours to complete.”
EMC is the ability of an electronic system to function as intended without negatively impacting the performance of other electronic systems. ISO and IEC compliance specifications include radiated (over-the-air transfer) and conducted (through the wiring harness) tests. Each covers emissions (emanating from the system outwards) and immunity (external sources affecting the subsystem under test) aspects of the electronic subsystem. This results in four distinct EMC tests, which are listed below with an example of an EMC test standard cited for each type of test:
1. Conducted immunity test (ISO11452-4)
2. Radiated immunity (ISO 11452-2)
3. Conducted emission (CISPR25)
4. Radiated emission (CISPR25)
A fifth test called electrostatic discharge (ESD) test is often discussed collectively with the above EMC tests. However, it is distinct, as it is performed by using an ESD gun to inject static charges onto the subsystem to measure its immunity. This article focuses on the four EMC tests in great depth.
Let’s look at each of the EMC tests in greater detail, using a capacitive sensing application as an example. We’ll examine custom EMC tests mandated by different automotive OEMs (such as Toyota, BMW, and VW) across geographical regions. We’ll also cover some system-level aspects of designing compliant electronic systems for the automotive environment and upcoming challenges in the way the industry performs EMC tests.
Overview of Defect Classes
EMC tests are performed to evaluate whether the system under test, as well as other systems in its vicinity, function flawlessly in the presence of noise. Some examples of EMC failure symptoms in an automotive environment are as follows:
1. Malfunctioning of noncritical electronic equipment (example: infotainment system, power window control, cabin lighting, etc.) when starting a car.
2. Malfunctioning of critical electronic equipment such as the engine control unit (ECU) or braking system due to a spike during a load dump condition.
3. Audible sound resembling the cranking of the engine or starter motor as an audio output from the infotainment system of the car.
4. Permanent damage to an electronic circuit during the jump start of a car.
5. Spurious and unintended reset of an electronic system due to an electrical disturbance.
Automotive manufacturers specify the severity of such symptoms using defect classification. A typical example of defect classification is described in Table 1.
It must be noted that the severity level is not an indication of the criticality of the defect. All systems, whether critical or not, can have defects in each defect class listed above. However, the acceptance level is different for critical systems such as ECU, braking control, airbag control, power train, and so on. Such systems must not have defects even in class 5. For non-critical systems such as multimedia, infotainment, power windows, and so on, the acceptance level is often lower. Some defects in class 4 and class 5 are usually allowable, depending on their overall effect on the system. The acceptance level is determined by the manufacturer, depending on the possible outcomes of the defects in each subsystem of the car and the cost involved in solving non-critical issues.
IC-Level Automotive Qualification
All electronic components, especially semiconductor ICs that are selected for use in an automotive embedded system, must be capable of surviving and functioning flawlessly in the harsh automotive environment. Device-level standards such as AEC-Q100 are laid out by the Automotive Electronics Council as a guideline for semiconductor manufacturers to qualify their devices for use in a car.
The following is a partial list of important parameters of an IC that are tested for AEC-Q100 qualification:
1. Part operating temperature test (typically up to 125 degrees Celsius)
2. Accelerated life-time simulation test
3. Stress test
4. Wire Bond and solder ball Sheer Test
5. Human Body Model (HBM) ESD test and latch-up test
6. Non-volatile memory endurance testing
7. Short circuit reliability characterization
Due to the high-reliability requirement for electronic components in an automotive system, the system designer must select ICs that are AEC-Q100 qualified. Because of the complexity involved, it is advisable to contact a manufacturer for specific information on its AEC-Q100 qualified components beyond the scope of this article. Manufacturers may also provide detailed literature and applications support for each of their products to pass automotive EMC tests.
System-Level EMC Test Methodology
EMC experiments quantitatively measure if the system under test is robust enough for automotive use. The system must also be “quiet” enough in its functioning such that it does not disturb other systems in the vicinity. For robustness or immunity tests, noise is injected from an external source using either a wiring harness (for conducted immunity) or RF antennas (for radiated immunity). System functionality is observed during the noise injection. The frequency of the injected noise is varied, and its intensity is controlled as defined in the standards specification.
For quantitatively measuring the conducted emissions from a system under test, the frequency spectrum of noise voltage appearing on the wiring harness is plotted against its spectral intensity. For radiated emissions, the spectrum of noise is received by an appropriate RF antenna placed at a distance that is predefined in the standards specification. In either case, the spectrum must be below the acceptable limits defined in the specification. Some commonly followed automotive EMC specifications are listed in Table 2.
In addition, automobile manufacturers often have proprietary specifications, many of which are more stringent than the generic standards. Some examples of proprietary automotive EMC standards are listed in Table 3.
Conducted Immunity Tests
For automotive environments, conducted immunity tests are the most severe among all the EMC tests, making it the most difficult to achieve compliance to immunity tests. Conducted immunity tests are performed by injecting noise onto the system through the wiring harness. The system should function flawlessly in the presence of the noise.
The block diagram in Figure 1 shows an example of a Cypress PSoC system that performs the function of a capacitive sensing user interface for controlling the HVAC control unit inside a passenger vehicle. In the case of the ISO11452-4 EMC standard test, an RF noise generator injects noise through an RF-coupled probe onto a wiring harness that supplies power to the system under test. The noise frequency is in the range of 1MHz to 400MHz, and the intensity is 200mA of loop current in the RF probe as shown in Figure 2.
Let us take the example of a capacitive sensing interface such as the Cypress CapSense interface. Capacitive sensing is quickly replacing mechanical buttons in a multitude of automotive user interfaces. The functional correctness for a capacitive touch-sensing system is defined by the absence of the following defects:
1. No false touches, i.e., no detection of a touch input by the system when a human touch is not present at the input sensor.
2. No reduction in touch sensitivity, i.e., the user should not have to press the input sensor with additional force for the system to detect a touch input.
3. No stuck state, i.e., the system should not get stuck in the touch-detected state after the user removes the touch input.
4. Intermittent reset of the system due to EMC noise injection is unacceptable.
For passing the conducted immunity test, it is important to design the system with great care, including the design of the voltage regulator circuit and the use of protection circuit elements such as transient voltage suppressor (TVS) diodes, etc. In particular, designing a low-ripple automotive power supply to supply a stable voltage to the subsystem from the 12V automotive battery is crucial for product compliance. The graphical plot in Figure 3 shows the output voltage ripple on the supply rail during an ISO11452-4 test when using a non-automotive power supply circuit (blue plot representing a Rev1 board) in comparison with a well-designed automotive power supply (red plot representing a Rev2 board).
The ripple is plotted as a function of the noise frequency injected during conducted immunity tests. As seen in the plot, the ripple in the 100MHz region is greatly reduced when the power supply circuit is well designed. This test was performed for a Cypress Semiconductor PSoC4 based capacitive sensing system designed for an automotive HVAC user interface control unit.
Another immunity test, the ISO7637-2 transient immunity test, is peculiar to automotive systems, since it emulates the typical electrical noise generated on the wiring harness during in-car events such as jump-start, load-dump, ignition cranking, reverse battery, etc. The use of appropriate protective devices such as TVS diodes and series inductors in the power path help reduce the effect of transients on the system under test.
Radiated Emission Tests
In an automotive environment, radiated emission tests are particularly important for high-frequency switching systems such as touch-screen infotainment systems and RF transceivers, including Bluetooth and key fob systems in the cabin. The radiated emission test is based on using an RF antenna to measure the spectrum of frequencies emitted from the system under test. Spectral intensity (FFT amplitude) is measured using a spectrum analyzer. Figure 4 shows a laboratory set-up for measuring radiated emissions in an anechoic chamber.
The frequencies of interest and their permission amplitudes limits are defined by the respective EMC standard. With recent automotive systems generating emissions with even higher amplitudes and frequencies, the EMC standards now include frequencies up to 5GHz. In most cases, a single RF antenna cannot be used to measure the emissions across a wide range of frequencies, hence multiple antennae are used through several bands of frequencies. This process involves equalizing the antenna impedance for all the antennas so that the results across the entire range of frequencies can be compared. Currently, OEMs are adopting a single wideband bow-tie antenna to cover the entire range of frequencies, as shown in Figure 5.
One of the ways to reduce emissions at higher frequencies is to control the rise time of the switching signals that cause the emissions. Another method is to sweep the frequencies being generated using spread spectrum techniques so that the overall energy is distributed across a wider band of frequencies. Sometimes these methods are integrated by manufacturers with specific components. For example, Cypress employs provide both these techniques in its capacitive sensing technology for effectively achieving compliance concerning radiated emissions.
Radiated Immunity Tests
Radiated immunity tests also use the same antenna as that used for emission tests. For the test, RF noise is injected onto the system under test through a radiation medium (over-the-air). The criteria for passing the tests is similar to that mentioned for conducted immunity tests earlier in this article. Methods such as shielding and metallic enclosures are used for reducing the impact of radiations on the system under test.
Conducted Emission Tests
Conducted emissions deal with the noise injected by the system under test onto the power lines (wiring harness) that often act as a common bus rail across multiple subsystems. In most cases, electronic subsystems in an automobile are not the primary source of electrical noise on the wiring harness. In cases where they significantly contribute to the conducted noise, results can be improved by designing the system with proper bypass capacitors at the power entry point.
Challenges in EMC Compliance Criteria
With electronic subsystems performing critical functions in modern automobiles, the compliance criteria have become more stringent with time. Many automotive OEMs are now mandating that even non-critical electronic subsystems must pass the EMC immunity tests without any defect, not even a defect level 5 as listed above. Also, automotive EMC testing is becoming more and more involved as each function of the subsystem is thoroughly tested at each frequency at which noise is injected. Non-critical characteristics such as user response time and low-power modes in subsystems, which are not even visible to the user if they deviate during EMC tests, are tested thoroughly, and no deviation is allowed.
This also brings out the need to automate the process of EMC testing since a typical EMC test now takes several hours to complete. For example, an exhaustive EMC test for a capacitive sensing subsystem can take over three hours for only the conducted immunity test and over 20 hours for the whole suite of EMC tests. With mobile and Internet-of-Things (IoT) applications likely to be placed in the vicinity of the system under test (i.e., the user is carrying a mobile phone or a Wi-Fi enabled device in the car), EMC tests are stretching the frequency band for radiated immunity and emission tests beyond 5GHz.
Most EMC tests are performed in an anechoic chamber where the walls are made of absorptive material that prevent reflections of the radiated emissions. This is an ideal set-up and far from the actual environment in which an automotive subsystem resides. Hence, some automotive OEMs also use a metallic chamber or a mode-tuned reverberation chamber where intentional reflections can be directed towards the system under test. This way they can study the effect these reflections have on the functionality and immunity of the system.
As can be seen, designing an automotive electronic subsystem is becoming even more difficult, especially from a compliance viewpoint. The added challenge of keeping cost low and minimizing the ever-increasing sources of radiation makes design even more complex. Sometimes, an end user may carry a non-compliant or non-automotive grade device into the car, appreciably affecting system functionality. What’s more, this effect is difficult to detect using standard EMC tests. A common example is the peculiar audio noise heard from the infotainment system just before a non-compliant mobile phone receives an incoming call in a car.
EMC Infrastructure Challenges
The equipment used for EMC tests is expensive. Often, a typical automotive Tier-I supplier cannot justify the cost of procuring the equipment for in-house use. A common practice is to rent the premises and equipment from a certified EMC lab. However, this involves considerable cost on an hourly basis. Also, if an EMC test fails, the design engineer needs to revise the system (hardware, firmware, etc.) and repeat the test at the lab at additional cost. With the numerous Tier-I suppliers waiting for time-slots at the limited number of EMC laboratories, there is usually a considerable delay from the time the test slot is requested to the time it becomes available to the designer, further impacting the overall time to market.
One of the ways to reduce the cost and inefficiencies of working with an external lab is to use pre-compliance equipment in-house at the design site to tune the system. Pre-compliance equipment does not provide accurate measurements of the intensity, but it does provide an approximate signature of the noise spectrum that the designer can use to fine-tune the system.
Product compliance and electromagnetic compatibility for automotive electronic subsystems has become increasingly challenging over time. It is important that electronic products are designed with a focus on compliance right from the design stage. Testing can be challenging, especially with limited in-house RF expertise and testing equipment. For this reason, using compliant silicon isn’t sufficient to ensure a company will pass EMC testing for automotive subsystems. Silicon manufacturers need to actively support their silicon. While application notes can describe best practices for achieving EMC compliance, each subsystem is unique. Thus, it is important that you have access to experienced application engineers who can help you review hardware schematics and PCB layout and tune firmware parameters.
Shantanu Prabhudesai works at Cypress Semiconductor, India as part of the global automotive applications team. His primary role at Cypress involves defining the system architecture for evaluation boards and validation platforms, including hardware-software co-design of embedded systems. His special interest is in electromagnetic compatibility (EMC) compliance for automotive end-systems, for which he is currently leading the EMC Center of Excellence at Cypress. As part of the Cypress applications team, he also engages with customers worldwide to help them achieve their EMC expectations with the use of Cypress products. Shantanu graduated in 2005 with a masters’ degree in electronics design from the Indian Institute of Science, Bangalore.
1. European Telecommunications Standards Institute (ETSI), http://www.etsi.org/
2. Automotive Component EMC Testing, www.ets-lindgren.com/pdf/automotivecomponentemctesting.pdf
3. Automotive Research Association of India, https://www.araiindia.com/
4. Automotive EMC Testing: CISPR 25, ISO 11452-2 and Equivalent Standards, http://incompliancemag.com/article/automotive-emc-testing-cispr-25-iso-11452-2-and-equivalent-standards/
5. Proceedings of 2012 ESA Workshop on Aerospace EMC 21‐23 May 2012, Venice, Italy. http://esaconferencebureau.com/docs/12a05_docs/programme-emc-27-april-2012.pdf?sfvrsn=2
6. Automotive Electronics Council – Component Technical Committee: http://www.aecouncil.com/AECDocuments.html
The author would like to acknowledge the information provided by two certified EMC labs, TUV Rheinland (Bangalore, India) and TUV SUV Automotive EMC lab (Detroit, Michigan, US) in terms of laboratory facilities and capabilities at their respective premises.