The design of modern electronic systems come with a unique set of challenges that can be solved by introducing galvanic isolators.
The design of modern electronic systems in applications as varied as industrial automation, power supplies for telecom base stations, and on-board chargers (OBC) for electric vehicles (EV) comes with a unique set of challenges:
- Providing safety from high voltages – for electronic control equipment as well as human operators
- Effective communication across sub-systems with relatively high ground potential differences
- Preventing electrical noise from corrupting sensitive signals
These challenges can be solved by introducing galvanic isolators in the circuit design. Galvanic isolators are devices that couple an electrical data or control signal across an insulation barrier without any flow of current through it, thereby enabling signals to be transferred while blocking noise. The insulation barrier also safeguards equipment and human operators from high voltages.
The earliest isolators were optically coupled devices, also called opto-isolators or optocouplers, “optos” for short. The first patents for optos were issued in the 1960s. Earliest forms consisted of miniature incandescent light-emitting bulbs on the primary side, a clear (optically transparent) plastic, which acted as the insulation or dielectric layer as well as the optical path, and a photo-resistor on the secondary side, whose resistance was modulated by the amount of light falling on it. Later developments introduced more sophisticated opto-coupled devices that made the system designer’s job a little easier.
They were all basically some sort of light-emitting structure (the miniature bulbs were replaced by a semiconductor-based light emitting diode or LED) on the primary side, coupled with various forms of light-sensitive devices, such as a photo-resistor, photo-transistor, photo-diode, or triac, that made the device suitable for a range of dc and ac applications. Optos were basically the only solution available until CMOS-based digital isolators were developed in the late 1990s using inductive (magnetic) or capacitive coupling to transfer the signal. Figure 1 highlights the technology differences between optocouplers and digital isolators.
Figure 1 Basic operation of opto vs capacitive coupled CMOS isolator
Figure 2 shows examples of X-ray images of an opto and a digital isolator to help visualize the physical construction of these devices.
Figure 2 X-ray images of an optocoupler assembly (left) and a digital isolator assembly (right)
A prominent characteristic of the opto is the aging problem. The quantum efficiency of the LED, defined as the total photons per electron of input current, decreases with time at a constant current. This is largely due to the electrical and thermal stressing of the PN junction. This had implications for the optocoupler’s long-term stability and operation, especially at high temperature operation. The designer can compensate for aging by doing several things:
- Decrease real operating life
- Decrease operating diode current and ambient temperature
- Avoid peak transient currents
Of course, these actions put limitations on the use case, since isolators are most useful in systems where such conditions exist intrinsically. Digital isolators have no such physical limitations. Since optos basically function by switching the PN junction diode, they have a relatively slow switching rate. Therefore, optos are only capable of lower data rates with large propagation delays and skew.
Industry trends and CMOS digital isolators
In an increasingly bandwidth- and power-hungry world, new CMOS-based digital isolators provide an ideal solution. The most common applications for isolation are in the industrial market – in equipment such as factory automation, process control, programmable logic controllers (PLC) or process automation controllers (PAC), inverters for motor control and uninterruptable power supplies (UPS). Industrial automation is the largest market for isolators, and designers of industrial systems value the high-temperature operation, superior part-to-part matching, low skew, and high noise immunity that CMOS isolators bring to the table. Other heavy-use applications include isolated power supplies used in telecom base stations and in servers that power the infrastructure behind our increasingly connected world – the Internet of Things.
The early adopters of digital technology were isolated power supply manufacturers. These power supplies were primarily used for servers and telecom base stations. For this market, the most critical parameter was power density, and the mantra was W/mm3. It helped that worldwide green initiatives for a cleaner environment also mandated higher efficiency to reduce wasted energy. As it turns out, having a higher efficiency system also implies less heat loss, which leads to further reductions in system size because space-consuming heat sinks were no longer required. The biggest impact of CMOS digital isolator technology was on the timing characteristics of these new isolator devices compared to optos.
Since these were not based on switching the PN junction of the LED for enabling signal transfer, the switching rates improved by an order of magnitude. Combined with the advantages offered by the smaller geometries and the more repeatable and stable manufacturing process used by standard CMOS silicon technology, timing parameters such as propagation delays, pulse width distortion or skew, part-to-part matching, and common mode transient immunity (CMTI) were vastly improved. In the isolation industry, CMTI basically refers to common mode noise rejection capability, measured as a voltage slew rate, kV/µs. The limitations of optos were due to the manufacturing process involving compound semiconductor technology, which is better suited for optical operation rather than for fast and accurate devices. The inherent advantages of digital isolators helped power supply OEMs tighten up the power converter control loop timing, thereby gaining efficiency.
Another fast-emerging market for digital isolators is automotive. Whereas traditional internal combustion engine (ICE) based automobiles hardly used any isolators, this has changed with the introduction of EVs. The various forms of hybrid electric vehicles (HEVs) and EVs have a high-voltage battery ranging from 200V to 400V today, and higher voltages are planned in the future to enable higher power and/or capacity for maximizing the distance per charge. This high-voltage battery necessitates the use of isolators for safety and signal transfer across different voltage domains within the vehicle. Virtually all major automobile manufacturers have an imminent EV/HEV rollout plan. The automotive industry is also proving to be an early adopter of digital isolator technology due to its superior high temperature operation, stability, and noise immunity. End applications such as battery management systems (BMS) and chargers are driving the need for isolators in the EV/HEV market.
[Continue reading on EDN US: Advantages of optos]
Ashish Gokhale is a senior product manager for Silicon Labs’ power products, focusing on the company’s digital isolation portfolio.
- Optocouplers in intrinsic safety applications
- Where do we start? Optocoupler or Digital Isolator?
- Power Tips #81: Make sure your optocoupler is properly biased
- Keeping FIT: Optocouplers vs. capacitive-based digital isolators
- Thermal characteristics of optocouplers
- Battery management system keeps electric engines safe
- Designing battery-management systems
- Understanding isolator standards and certification to meet safety requirements