September 22, 2020


Technical support

Decades ago, electronic engine control transformed automobiles from purely mechanical systems into a sophisticated combination...

Decades ago, electronic engine control transformed automobiles from purely mechanical systems into a sophisticated combination of sensors, microcontrollers, and other electronics. Automotive designers quickly embraced this new technology and expanded their expertise to include electronics. With the advent of electric vehicles and hybrid electric vehicles, history is repeating itself. Leaving behind the low voltage, low power 12V systems, modern EV and hybrid systems harness the power of 800 V systems, with even higher voltages on the horizon.

Much like learning electronics decades ago, automotive designers now must learn and adopt a new technology: galvanic isolation. Galvanic isolation separates the high voltage from the low voltage domains in a vehicle and is critical to ensuring the safety and operation of modern, high voltage vehicle systems. The technology behind galvanic isolation itself has also advanced rapidly. The days of the simple optocoupler are fading as advances in CMOS processes have opened the door to cutting edge, digital isolation. This new class of isolators presents many benefits and a few new challenges. Fortunately, the most common isolation challenges an automotive designer faces can be overcome with a little background knowledge and some simple design techniques.

The problem with industrial grade
Before diving into isolation, some background on automotive semiconductor devices is required. When designing any automotive system, automotive grade devices must be used. Many industrial grade devices include an AEC-Q100 qualification, however an AEC-Q100 qualification is usually not enough to meet the stringent demands of today’s vehicles. This is especially true of the isolation devices, as they are a critical component from both a safety and system operation perspective. Often isolators relay control signals, measure key voltages and currents, and safely communicate with the main vehicle control unit. If the isolator fails, the entire system will fail.

Isolation creates a very high impedance between two grounds in a system, effectively eliminating the electrical connection between them. (Source: Silicon Labs)


True automotive grade products leverage special foundry monitoring, control, and methodologies to drive to zero defects (DPPM) in addition to having passed an AEC-Q100 qualification. Detailed product information reports, including Production Part Approval Process (PPAP) and International Material Data System (IMDS) documents, and priority failure analysis is also included with most automotive grade material. These additional steps and documents ensure the isolator, or any semiconductor device, is truly ready for use in a vehicle. 

A primer on galvanic isolation
The 12V automotive systems of the past did not require isolation, yet isolation is now key to high voltage electric vehicle systems. For many automotive designers, isolation isn’t an unknown concept but one they are less familiar with. Simply put, galvanic isolation separates two circuits with a high impedance barrier that prevents the flow of current between them. Isolation devices are used to allow communication across the isolation barrier. For example, in a traction inverter the system controller might be in a low voltage power domain, while the field effect transistors (FETs) driving the motor are all in a high voltage power domain.

Using an isolated gate driver, the system controller can safely control switching of the FETs from the low voltage realm. In addition, isolation provides critical safety to the system by simultaneously protecting the low voltage circuits from the hazardous high voltages. Beyond safety, the isolation barrier greatly reduces the noise caused by ground loops between two different circuits. This is especially useful in vehicle communication systems, where data integrity is key. With an understanding of how isolation is used and why it is needed, safety requirements are the first challenge most automotive designers face.

A few things that can safely be said about safety
With any high voltage system, it is absolutely essential that the designer consult with their internal safety team and/or external safety compliance agency to ensure that their system complies to its standard’s safety requirements.

The major role of isolation in a vehicle is to ensure system safety. Thus, automotive designers must understand the key safety features of an isolator and its implications. Beyond the isolator’s voltage rating, there are typically three common topics that cause designers confusion: creepage, clearance, and safety agency marks.

Figure 2: Creepage and clearance graphically illustrated. (Source: Silicon Labs)

Creepage and clearance address the prevention of currents inadvertently flowing around the barrier contained in the isolation device, while the safety agency marks help validate the ability of the isolation barrier within the device to prevent the flow of current through the isolation device. Figure 2 illustrates the difference between creepage and clearance. Creepage is the shortest distance around the package between two pins on either side of the isolation barrier.


Clearance is the distance between two pins on the package representing the shortest line of sight distance (in air) between the two isolated sides. End system standards (e.g. IEC 62368-1 for Information Technology) all have minimum creepage and clearance distance requirements for a given high voltage and should be consulted before starting a design. If these distances cannot be met in air, then conformal coating is required, adding a host of issues including costs and rework difficulties. The designer should also pay special attention to the package of the isolator, as some older packages have residual tie bars that are exposed on the edges and reduce the creepage rating of the package, as shown in Figure 3. After selecting a device with sufficient creepage and clearance, the designer must now review the safety approvals of the isolation device.

Figure 3: Exposed tie bars (circled) reduce the amount of creepage. (Source: Silicon Labs)

For many designers, safety certification can quickly become a headache. Understanding what safety marks represent and how they fit with system certification makes the process less painful. A safety agency’s mark on a product, such as VDE, UL or CSA, means two key things. First, it means the product has passed the testing requirements for that particular mark and has a test certificate and test report from the safety agency. Second, and this is often missed, the safety agency regularly audits the production facility to ensure all devices are built the same as what was tested in the safety report. Without this step, it is possible for a device to have a safety test report but for the manufacturer to change how it is built later on and no longer pass the safety tests. Safety agencies will only provide their mark if the device has satisfied both requirements.

Why does this matter to an automotive system designer? It matters because isolation devices are typically considered safety critical by safety agencies. Using isolation devices with the appropriate safety marks will greatly simplify certifying the entire system. Most isolation manufacturers make their safety certificates readily available on their websites or through their sales teams. If they don’t, that’s a big red flag. With safety certifications in hand and the appropriate marks on the isolation device, designers can move to something they are more familiar with, resolving device failures.

The most common failure mechanism
What is the most common reason an isolator fails in a vehicle? No, the answer is not the isolation barrier failing over time. In fact, the isolation barrier in a digital isolator rarely fails. Electrical overstress (EOS) is by far the most common reason an isolator fails in an automotive application. EOS can be caused by a wide variety of issues depending on the application and type of isolator being used. Two common issues are exceeding power supply and I/O voltage ratings and issues surrounding ESD.

Like any semiconductor device, isolators devices will specify “absolute max” values for power supply and input voltage ranges and the device can be damaged if these values are exceeded. Most digital isolation devices require power on both sides of the isolation barrier, VDDA and VDDB. For digital isolators, that pass digital I/O signals across the isolation barrier, the supply voltages for each side of the isolator have an input range typical for CMOS devices. Likewise, the I/O pins need to stay in the range of VDD and can rarely tolerate anything over VDD.

However, isolated gate drivers often offer a much wider VDD range on the side of the isolator driving the power device (FET or IGBT), as many power devices need gate voltages higher than what typical CMOS can provide. See Table 1 for a comparison of the supply and I/O voltages for Silicon Labs’ Si86xx digital isolators and Silicon Labs’ Si823x isolated gate drivers. For the supply rails, a simple and common fix is to carefully follow the manufacturer’s recommendations for proper bypass capacitors and place them as close as possible to the device.

digital isolation table
Table 1: Digital isolator and isolated gate driver supply and I/O voltage ratings. Click the image to enlarge. (Source: Silicon Labs)

ESD events can be much more complex and difficult to trace down. For automotive applications, it is important to understand what the AEC-Q100 ESD tests are evaluating and what type of ESD the isolator will experience in the system. For components such as isolators, AEC-Q100-002 tests human body model (HBM) ESD and AEC-Q100-011 tests capacitive discharge model (CDM) ESD.

As the name implies, HBM ESD is designed to test how robust a device is when a person with an electrostatic charge (e.g. from shuffling feet across carpet) touches the device. Figure 4 shows the test setup. CDM is quite different. In this model the device has the charge and dissipates that charge through a single pin over low impedance, metal to metal contact. Figure 5 shows the typical test setup for CDM. For automotive applications, the AEC standard tests CDM on each pin three times in the positive and three times in the negative direction. Understanding these two types of ESD tests allows a system designer to understand the ESD ratings for an isolation device and translate that into their system ESD requirements.

Figure 4: Typical test setup for human body model ESD testing. (Source: Silicon Labs) (Click image to enlarge it.)

Many system level ESD tests are done at much higher levels than device ratings and are tested at multiple points with very different paths to the pins of the isolation device. If the isolation device is the failing component in system ESD tests there are a few common fixes. If the ESD damage is occurring on only one side of the isolator, then often increasing the VDD bypass capacitor can help resolve the issue. Improved ESD protection for that side of the isolator may also be required, such as an ESD diode, RC circuit, or ferrite bead.

If ESD damage is occurring across the isolation barrier, with damage to both sides of the isolator, then simply adding a capacitor across the isolation barrier, between the two ground planes, may solve the issue. This is commonly called a Y capacitor and it must be rated for the same or greater isolation voltage as the isolator. Just like the isolation device, the Y capacitor also needs to be automotive grade.

Although DC current will not flow through a Y capacitor placed across the isolation barrier, it may allow a small increase in the amount of AC “leakage current” that flows between the isolated power domains. The value of the Y capacitor must be limited to ensure the AC leakage current requirements of the application are not violated. If a Y capacitor does not solve the issue, then likely an isolation device with a higher surge rating is required.

Figure 5: Typical test setup for capacitive discharge model EST testing. (Source: Silicon Labs)

The need for isolation in vehicles is relatively new, as the 12V systems of old had little need for higher voltages and high-power electronics. Electric vehicles and hybrids have changed the industry demands. The latest automotive technology focuses on complex electrical systems that operate at higher voltages than ever before. This has compelled the automotive industry to adopt digital isolation technology and made it essential for their future success. Bringing new technology to market, particularly the automotive market, always offers challenges. However, today’s digital isolation technology makes adding isolation and controlling complex electrical systems simpler than ever. As the automotive market and governments drive the electrification of vehicles, automotive designers will continue to rely on the safety, reliability, and functionality of isolation devices to deliver cutting edge systems.

Charlie Ice is a senior product manager at Silicon Labs. He is focused on the company’s Power over Ethernet (PoE) product line. Charlie joined Silicon Labs in 2018 with more than 10 years managing products in the technology industry. Charlie holds a Master and Bachelor of Science in Electrical Engineering, both from Rice University in Houston, Texas.



The post Digital Isolation: What Every Automotive Designer Needs to Know appeared first on EETimes.

Source link