Delivering Safety & Security Through Digital Isolation

Author : Jian YE, Product Line Director, NOVOSENSE Microelectronics

06 March 2024

Figure 1: Operation of electromechanical relay
Figure 1: Operation of electromechanical relay

Silicon integration has revolutionised many different sectors and is continuing to do so in places where there is a need to handle electrical energy at scale. Automotive, industrial and green energy systems all call for the ability to use power transistors - so as to capture energy or convert it into motion through electric motors. This article looks at ensuring the protection of the delicate accompanying devices.

Much of the progression is thanks to access to control techniques for motors and actuators that would have been unthinkable a couple of decades ago. Unfortunately, the highly integrated devices needed to implement these sophisticated control algorithms are also very sensitive to the high voltages and currents required to drive motors or convert energy. Even a short-lived 100V spike can easily cause permanent damage to advanced semiconductor devices - and spikes of this magnitude or greater are commonplace, because high-voltage systems are prone to electro-static discharge (ESD), switching pulses and power disturbances. If these surges are allowed to pass from motors and power supplies into other parts of the system, they can cause safety issues, such as electric shocks to the human body, or they can potentially start fires. High-voltage devices are fabricated using materials and design techniques that allow them to handle these surges or block them without internal damage. 

A history of isolation
Because of differences in construction, isolation between the high-voltage/high-current sections of circuitry and more sensitive electronics is a necessity. In traditional hardware designs, the typical mechanism for providing control over the high-voltage electrical circuitry from a low-voltage subsystem was via an electromechanical relay. First conceived in the mid-1830s, by scientist Joseph Henry, such relays’ behaviour relies on electromagnetic induction and the ability to pass a signal from one voltage domain to another without direct electrical contact. In his experiments, Henry used an electromagnet to control the opening and closing of a high-current electrical path - with a selectively powered wound coil moving a switch element. 

This structure became the basis for many simple electrical systems that need to control power passing to a high-voltage, high-current domain. However, the relay is slow-acting and also prone to electrical arcs. Repeated damage from arcs generated during switching will result in the relay’s operational life being reduced substantially. For these reasons, the relay is best suited to situations where the switching is at a very low frequency. 

A continuing application for relays is of course in reaction to safety - cutting off power to a high-voltage circuit if sensors on the low-voltage logic board detect a problem. However, to allow better real-time control of power circuits, engineers developed the optocoupler. This pairs a photodetector with an accompanying light source, which is typically a light-emitting diode (LED). 

Figure 2: Optocoupler operation
Figure 2: Optocoupler operation

The gap between the 2 halves of the optocoupler device provides the key to its isolation properties, as there is no direct electrical path connecting them. An optocoupler-based design can easily meet the creepage requirements mandated by many safety standards as long as the isolation distance inside these devices is large enough. In operation, a photodetector measures the light intensity of the LED in real time by converting the absorbed photon energy into an electrical current that is directly proportional to the transmitted light level.

The assembly is smaller than a relay and provides a much faster response to changes in the control signal. Thanks to these attributes, optocouplers have made it possible to develop subsystems such as switched-mode power supply (SMPS) units and speed/torque control motors implementations.

Challenges facing optocouplers
Continuing pressure on system size, responsiveness and energy efficiency presents challenges to optocoupler-based designs. Optocouplers are typically single-channel devices, which leads to high cost and board area in situations where many control signals need to pass between the low- and high-voltage regions of the PCB. Though the power draw of the LED and photodetector may be much smaller than that of a motor or high-voltage power supply, the consumption is continuous, which may not be appropriate for circuitry that is expected to be in a low-power mode much of the time, such as metering. 

Typically, the maximum effective data rate an optocoupler can support is in the range of 50Mbits/s, and the cost is increased when the speed is higher. As control strategies become more sophisticated and make use of devices with very high switching speeds, such as wide-bandgap (WBG) semiconductors, the response time of the optocoupler becomes an issue. Another area where the optocoupler is becoming an obstacle to circuit design is in distributed control - where high-speed networking is used to send commands and data to different parts of the system. This is a key element of the architectures that are now being used in smart meter design. 

Due to the need to realise intelligent control, the architecture of the management unit and the metering unit need to be isolated from each other - the metering unit belongs to the high-voltage domain, while the management unit is a low-voltage domain element that may have contact with a user. Between these 2 units, high-speed SPI communication is needed to meet the demand for data transmission. In this case, the optocoupler is limited in its speed, propagation delay and level of integration, which cannot meet the system requirements. This is where digital isolation comes in.

Figure 3: Block diagram of capacitive isolator product and schematic diagram of OOK modulation
Figure 3: Block diagram of capacitive isolator product and schematic diagram of OOK modulation

Digital isolation techniques
There are 2 primary forms of digital isolation. The first is similar in some respects to the relay because it uses electromagnetic induction to pass a signal from one side of an isolation barrier to the other, though the action is much faster. The other is capacitive isolation, which takes advantage of the coupling effects encountered with electrical fields to pass signals across an insulator. High-frequency signals couple easily across the 2 sides of a capacitive isolator while blocking any direct electrical currents.

Though the core principle is effective, a number of design and manufacturing techniques are needed to guarantee protection against high-voltage spikes. For example, the construction and shape of the capacitive elements are important in determining the voltage withstand capabilities of the isolator. This is because they factor into the electric field distribution of the elements within the device. By paying attention to these proprieties, capacitive isolators can easily meet requirements for protection against electric shock and damage to circuitry from spikes and surges with maximum voltages of 5kV and beyond.

An important question is how the signal being passed across the isolation barrier should be modulated. One technique that has proven to be highly reliable and successful in practice is on-off keying (OOK). Using this form of modulation, a digital ‘1’ is represented by the presence of a high-frequency signal, which will be detected across the capacitive barrier. A digital ‘0’ is the absence of this signal. This readily supports digital values and the pulse width modulation (PWM) signals. Because of the rapid response of capacitor plates to changing electrical fields, OOK can easily support data rates up to 150Mbits/s.

Because other electric fields within the system can couple with the capacitive plates, common-mode noise may be a concern. Smart design techniques can suppress the effects of this noise and avoid interference being injected into the receiver’s signals. One way, for example, is to adapt the OOK modulation scheme. 

Adaptive OOK technology uses a detection circuit inside the IC to analyse the signal on both sides of the modulation barrier - in order to determine the degree of common-mode noise to which the device is being subjected. Only when the instantaneous interference levels are relatively high does the circuit need to perform more functions to suppress interference. In many scenarios, common-mode interference is short-lived. 

Figure 4: Circuit implementing capacitive isolation IC using adaptive OOK
Figure 4: Circuit implementing capacitive isolation IC using adaptive OOK

This adaptive approach strikes a better balance between power consumption and robustness than techniques based on static common-mode noise suppression. The technology involved is applied by the NOVOSENSE NSI824xWx series of industrial-/automotive-grade capacitive isolators. These devices protect against voltages up to 8kVrms and provide 200V/µs of common-mode transient immunity while supporting communication speeds as high as 150Mbits/s. 

Delivering adequate isolation is a critical design requirement for each and every electronic product. Pressures to meet the performance, power and size requirements of ever-more complex systems means that conventional optically-based isolation techniques may no longer be appropriate - thereby forcing engineers to seek out new approaches to protection and safety. Based on adaptive OOK technologies, the latest cost-effective, compact and high-performance capacitive isolator technologies, such as those made by NOVOSENSE, are a very effective choice for markets such as automotive, industrial control and green energy, where isolator usage needs to address speed, voltage and form factor challenges while taking account of differences in safety standards.

Contact Details and Archive...

Print this page | E-mail this page