Galvanic Isolation Solutions for Industrial Automation Applications

Levels of Isolation

There are four distinct levels of isolation: 
- When isolation is used to enable the system to function properly, but not necessarily to serve as a barrier against shock, it is called functional isolation. 
- Where isolation provides sufficient protection against electrical shock as long as the insulation barrier is intact, it is called basic isolation. In basic digital isolation we offer devices in single, dual, triple and quad channels, with isolation ratings up to 3kVrms. 
- Safety regulations require basic isolation to be supplemented with a secondary isolation barrier for redundancy, so that the additional barrier provides shock protection, even if the first barrier fails. This is called double isolation. 
- To make systems compact and save cost, it is desirable to have only one level of isolation that has the required electrical strength, reliability and shock protection of two levels of basic isolation. This is called reinforced isolation. When it comes to reinforced isolation, our portfolio includes dual and quad devices with speeds up to 100Mbps and isolation ratings up to 5.7kVrms, the industry's highest.

Understanding these definitions and their relevance to real applications allows you to pick the right isolator for your design. The isolation level and their testing methodology are discussed at length in the ‘High-voltage reinforced isolation: Definitions and test methodologies’ article from TI.

Some Key Isolation Parameters

The high-voltage isolation performance of an isolator is described by several key parameters. The most important are: 
- Maximum transient isolation voltage (VIOTM) and isolation withstand voltage (VISO) indicate an isolator’s ability to withstand temporary (less than 60 seconds) high voltage.
- Maximum repetitive peak voltage (VIORM) and working voltage (VIOWM) indicate the continuous voltage that the isolator can withstand throughout its lifetime.
- Maximum surge isolation voltage (VIOSM) indicates the maximum impulse voltage (waveform with 1.2-μs rise and 50-μs decay time) that the isolator can withstand.
- Creepage and Clearance: Distance along the surface of the package and through the air between pins on one side of the isolator to the pins on the other side. System level standards mandate minimum values of these parameters based on the working voltage, the peak transient voltage, and the surge voltage.
- Comparative Tracking Index (CTI) indicates the ability of the package mold compound to handle steady high voltage without surface degradation. A higher CTI allows the use of smaller packages for the same working voltage.

There are many other specifications in isolation datasheet, related to timing, power consumption, immunity to transients, and more. 

Capacitive Isolation Overview

TI's digital isolators use internal high-voltage silicon dioxide (SiO2) capacitors to form the isolation barrier. Internally the isolator consists of two chips connected by bond wires: a transmitter; and a receiver chip which contains the high-voltage capacitors.

Figure 1: On-Off Keying architecture used in TI's ISO78xx reinforced digital isolator family (source: TI)

Figure 1 shows the conceptual block diagram of one channel of a digital capacitive isolator (DCI). The ISO78xx family uses an advanced architecture called On-Off Keying (OOK) to transmit the signal across the silicon dioxide-based isolation barrier. The incoming digital bit stream is modulated with an internal spread spectrum oscillator clock to generate OOK signaling such that one of the input states is represented by transmission of a carrier frequency, and the other state by no transmission. The ISO78xx devices incorporate advanced circuit techniques to maximize the Common Mode Transient Immunity (CMTI) performance and minimize the radiated emissions due to the high frequency carrier and I/O buffer switching. 

Figure 2: Example isolator in a 16-pin package (source: TI)

A pin diagram of a typical digital isolator is shown in Figure 2. It consists of two supplies VCC1 and VCC2 , two grounds, GND1 and GND2, and input and output pins on either side referred to the respective grounds; pins 1 through 8 are referred to GND1 and pins 9 through 16 are referred to GND2. The output pins on either side are enabled if their respective ENx pin is high or open; otherwise the outputs are in high-impedance state.

We've put together a design guide to help you begin designing with TI's broad portfolio of digital isolators and isolated functions in the shortest time possible. Our portfolio includes the ISO78xx family of 5.7-kVrms reinforced digital isolators, the ISO73xx family of 3-kVrms digital isolators, and the ISO71xx family of 2.5-kVrms digital isolators, among others. 

Capacitive vs. Optical Isolation

Let's take a look at how capacitive isolation technology compares to the other gorilla in the room, optical isolation, which has been widely used in the past. 

Optical isolators, or optocouplers, achieve isolation by converting digital data into light pulses using an LED and then transfer information via a closed optical channel to a phototransistor or photo diode, which converts it back into current. The isolation is provided by the physical separation of the transmitter and receiver.

How do optical and capacitive isolation stack up? Let's compare the two in a couple of important areas.

Isolation Comparison

A TI capacitive isolator, with its SiO2 capacitors as inter-level dielectric, has two benefits. First, it's one of the most robust isolation materials with the least aging effects and, therefore, extends the life time expectancy of capacitive isolators well beyond those of competing technologies. Second, SiO2 can be processed using standard semiconductor manufacturing, thus contributing to significantly lower production costs. 

An optocoupler, on the other hand, gets its isolation at the packaging level. The LED and photocoupler are attached to a split lead frame separated by a physical gap of between 80 and 1000 microns, and a transparent insulating shield or silicone, relyin on a combination of a physical gap, polyimide tape, silicon filler and plastic mold compound for insulation. This hybrid construction results in greater part-to-part variation and complexity, which increases cost and lowers reliability.

Reliability Comparison

Semiconductor manufacturers and their customers are fanatical about quality and reliability. Highly sophisticated (and expensive) design, qualification and testing procedures are employed to make sure defective products are screened out before leaving the factory, and that once products are operating in the application, their reliability is the highest it can be. A failure in the field sets off a rigorous process of fault analysis and isolation, followed by corrective action to make sure the same failure can never happen again.

Failure in Time (FIT) - the number of failures experienced within a certain time period – is a common metric used by quality and reliability engineers. For semiconductors, it is typically defined as the number of failures that can be expected in one billion (109) device-hours of operation. 

Empirical data indicates a correlation between failure rate and temperature or voltage increases. For instance, failure rates at 40°C are three times lower than those at 55°C and 80 times better than those at 125°C; at 40°C using 20% lower voltage improves the failure rate by a factor of eight.

Opto-couplers experience a considerably higher failure rate than capacitive isolators. A comparison of FIT values for two different devices - one a digital isolator from TI and the other from a prominent opto-coupler company – yields the following table:

Figure 3: A comparison of the FIT data at 55°C for a capacitive-based isolator versus an optocoupler at 60% confidence

A Word About Electromagnetic Compatibility 

Before we discuss some real-world applications, let's briefly talk about electromagnetic compatibility (EMC).

It could be argued that experts in EMC relish its reputation as something of a black art – move a capacitor over here, tweak a layout over there, and problems mysteriously vanish... or get worse. Nonetheless, an appreciation of the relevant EMC compliance tests is critical to understanding how a digital isolator will perform in a real-world application. 

People have spent their lives on this topic, but never fear: to help you out, we've put together a comprehensive white paper ‘Understanding Electromagnetic Compliance Tests in Digital Isolators’ discussing the basics of EMC, the applicable standards, some pitfalls to avoid, and some common test setups.

Industrial Automation Applications for Galvanic Isolators 

There are numerous applications for isolation devices in industrial automation. In this section, we'll take a look at three different areas where isolation is used: analog data acquisition, high-speed digital communications, and isolated power.

Isolation Application: Analog Data Acquisition

There are many industrial applications - remote sensing or motor control, for example - where low-level analog signals must be detected, amplified and digitized in the presence of potentially dangerous voltages. An isolated analog-to-digital converter (ADC) is tailor-made for such situation. 

There are many different ADC architectures – successive approximation (SAR), pipeline, integrating – but for lower-speed use, the delta-sigma architecture combines high resolution with low power and low cost, making it ideal for many industrial sensing applications.

The AMC1304 is a isolated, precision delta-sigma (ΔΣ) modulator with an integrated LDO regulator. The output is isolated from the input by a capacitive double isolation barrier certified to provide reinforced isolation of up to 7000 VPEAK according to VDE V 0884-10, UL1577 and CSA standards. Used in conjunction with isolated power supplies, the device prevents noise currents on a high common-mode voltage line from entering the local system ground and interfering with or damaging low voltage circuitry. 

Figure 4: functional block diagram of the AMC1304 (source: TI)

The AMC1304 front-end circuitry contains a differential amplifier and sampling stage, followed by a second-order, switched-capacitor, feed-forward ΔΣ modulator. The AMC3104 can achieve 16 bits of resolution with a dynamic range of 81 dB (13.2 ENOB) at a data rate of 78 kSPS. 

Isolation Application: High-speed Digital Communication

High-speed digital communications are a must in industrial automation. A modern connected factory has a large number of machines, industrial robots, individual sensors, actuators, and valves under the control of embedded microcontrollers, industrial PCs and programmable logic controllers (PLCs). These devices talk to each other via a high-speed communication network with cable lengths hundreds, or even thousands, of feet long. The ground levels at each end of a thousand-foot Ethernet cable may be at very different levels, especially given the high voltages and currents drawn by large industrial machines such as smelters and motors. If there is an imbalance in these currents, the return current may be very large, resulting in a large differential voltage; this can cause damage to high-speed digital networks. 

You can avoid this situation by using a high-speed digital isolator in the network. TI offers a number of high-speed digital isolators for both specialized network applications and general I/O use. 

The ISO1176 is an isolated RS-485 differential line transceiver that operates up to 40Mbps. It is designed for use in PROFIBUS applications in industrial automation applications such as networked sensors and motor & motion control. 

The ISO1050 is designed for use with CAN, another popular industrial protocol. This device meets the requirements of the ISO11898-2 CAN specification and provides galvanic isolation up to 5000VRMS.

For high-speed digital I/O, the ISO7842 is a high-performance, quad-channel digital isolator with 8000 VPEAK isolation voltage and supports a signaling rate up to 100Mbps. It provides high electromagnetic immunity and low emissions at low power consumption, while isolating CMOS or LVCMOS digital I/Os with a VDE 0884-10 peak voltage rating of 4000V. 

Isolation Application: Flyback Power Supply

An isolated device requires that the power supplies on each side of the isolation barrier are also isolated from each other. For low-voltage low-power applications (1 or 2W), TI offers isolated dc-dc converters such as the DCH family that are integrated into a single package; for higher voltages and currents, though dc-dc converter designs typically use an integrated controller and discrete components, including a transformer for input-output isolation. 

Another power conversion application where isolation is a requirement and safety is the overriding concern is in ac-dc conversion. Again, a transformer provides isolation between potentially lethal voltages on the primary side and the rest of the system. 

Figure 5: UCC2891x flyback switcher (source: TI)

Depending on the topology, the control circuitry may be on either the primary or secondary side. Figure 5 shows an isolated flyback switching power supply using the UCC289x controller for primary-side regulation. The UCC28910 and UCC28911 are high-voltage devices with a 700V power FET that provide output voltage and current regulation without the use of an opto-coupler. A full circuit analysis can be found here, but the main circuit blocks are:

1. Input fuse resistor: limits the inrush current on the input capacitor when the line voltage is applied, and disconnects the line in case of input overcurrent.
2. Bridge diode: rectifies the input ac line voltage.
3. Line filter (L1, L2, R1, and R2): reduces EMI generated by switching.
4. Bulk capacitor (capacitors C1 and C2): stores the energy and reduces input voltage ripple.
5. VS divider resistors: determine output voltage regulation point; RS1 also sets the continuous switchover point. 
6. Drain voltage clamp circuit: protects power FET, and dampens oscillation due to transformer primary leakage inductance.
7. Transformer: provides isolation; auxiliary winding provides controller power.
8. Output stage: size of output capacitor COUT determined by desired no-load transient response. Pre-load resistor RPRL stops OVP tripping if no external load is connected.

In a push-pull converter, the transformer primary is supplied with current from the input line by pairs of drivers in a symmetrical push-pull circuit. The drivers are alternately switched on and off, periodically reversing the current in the transformer, so current is drawn from the line during both halves of the switching cycle. Push–pull converters have steadier input current than buck-boost converters, create less noise on the input line, and are more efficient in higher power applications.