Analog Isolation Techniques for Circuit Protection

ESD Protection Circuit: Analog Isolators

Circuit protection is the mechanism which protects circuits from damaging events like electrostatic discharge (ESD) and interference. Implementing circuit protection can be done using a wide variety of methods, including current limiting resistors, Zener diodes, and fuses. However, some applications may deal with highly sensitive applications that can require electrical isolation. One common example is the electrocardiogram.

The electrocardiogram is unique in that it uses conductive pads connected to a patient to measure the electrical pulses generated by the heart as it beats. Unlike optical heart rate methods, the electrocardiogram can provide significantly more information, including unusual heart patterns and weak beats. But attaching electrically conductive electrodes across a patient’s chest can be potentially fatal during events such as an electrical surge, where the surge finds its way across the patient and therefore causes cardiac arrest. For instances such as this, circuitry requires isolation, whereby the electrodes attached to the patient are electrically isolated from the mains powered electrocardiogram. The result is that any surge experience by the electrocardiogram cannot be transferred to the electrodes as the electrodes are electrically isolated (that is, there is no direct electrical connection).

This type of isolation is very easy to achieve in the digital world with devices such as optoisolators, but the analog world can provide challenges. Where the digital world consists of only 1’s and 0’s which are easy to transfer non-electrically (presence / no presence of light), analog signals have an infinite range whose value needs to be preserved.

So, how can an analog signal be isolated and what advantages / disadvantages are typically associated with all the many isolation solutions out there?

Isolation Transformer

One method that can provide some form of analog isolation is the transformer. Two coils that are magnetically coupled via an iron core can transfer energy to each other via the magnetic field. While these two coils are electrically isolated from each other, this type of isolation has its issues. The first is that while the two are electrically isolated, surge energy can be transmitted across the transformer. Since transformers often have a large inductance, they will be able to resist sudden energy spikes (such as those found from electrostatic discharge or ESD), but mains surges may be harder to resist (as they are in the low-frequency domain).

Another issue with transformers is that they operate in the AC realm, and since energy transfer only occurs during a change in magnetic field strength, DC analog voltages cannot be isolated using this method. What’s more, transformers are often designed to work at specific frequencies. This makes them all but impractical for analog measurements of signals with an unknown frequency.

Switching to Digital Signal

As stated previously, digital signals are very easy to isolate and analog signals can take advantage of this solution for isolation. An analog signal is first converted into a digital pulse width modulation (PWM)  signal, whereby the duty cycle of the PWM represents the analog value. This PWM signal, being digital, is then isolated using an optoisolator and the output of the optoisolator is then converted back into an analog signal using an RC circuit. This method provides protection from both electrical surges and ESDs as well as preserving DC values, thereby making it a more ideal choice over a transformer. However, converting an analog signal to a PWM signal can be problematic depending on how this is achieved. If the signal is converted using all analog circuitry (a triangular wave compared to the analog signal), then the infinite range of possible values is preserved. But if the PWM is generated using counters (or any computation device), then the analog signal is quantised. This quantisation means that the resultant isolated analog signal will not be true to the original signal and the output analog value will have a finite resolution with a level of uncertainty. The other major disadvantage of this method is that the circuitry used to convert the original analog signal is not isolated and so is prone to damage from ESD and other high energy sources.

Linear Optical Isolation

Linear Optical Isolation is a technique which uses an optoisolator in its linear range. While optoisolators are mainly designed for digital isolation, they are constructed from analog parts, including an LED (typically an IR LED) and a phototransistor.

If an optoisolator is used in the current domain (as opposed to the voltage domain), the output transistor has a linear relationship to the input current of the IR LED. However, this method is very problematic for a wide number of reasons. The first, and perhaps most important, is the fact that optoisolators have a very narrow range of linearity (to the point where they are essentially non-linear). Secondly, analog voltages need to be converted into a current (as opposed to a voltage), which requires op-amp circuitry. Thirdly, optoisolators from the same family will not all have the same characteristics, and therefore may behave differently, making them impractical for production goods (unless trimming circuits are included). To solve this issue, one special range of optoisolators exist, which includes two matched IR LEDs that can be used together in an op-amp circuit so that the non-linearity is fed back into the amplifier and therefore preserving the analog value. While this method of isolation provides true isolation, it is a difficult method to implement.

Isolation Amplifier

Isolation amplifiers are integrated circuits that employ one of the aforementioned methods for performing analog isolation and will be the most likely choice for engineers who require analog isolation. The reasons for using an Isolation Amplifier over a custom isolation circuit include that the manufacturer has already designed the complex circuitry to produce a linear relation (if using the linear optoisolator method), and that the entire solution fits onto a signal chip. Some isolation amplifiers use an internal transformer for analog isolation. This method of isolation is achieved by using a voltage-to-frequency converter to convert an incoming analog voltage to a carrier wave with a specific frequency, passing this carrier wave across an internal transformer, and then reconverting the carrier wave into the recorded analog voltage using a frequency-to-voltage converter.

Mechanical Methods for Analog Isolation

Some scenarios can call for more creative and “out-there” solutions. Mechanical methods can be used for analog isolation, which may be ideal in scenarios involving very high energy. One example of mechanical isolation would be an electrically controlled potentiometer, whereby a motor (more specifically, a servo) controls the position on a potentiometer. Depending on the read analog value, the servo rotates the potentiometer to adjust its rotation angle to correspond to an analog voltage (if used as a potential divider). While methods such as these are very slow and not ideal for high frequency analog signals, they can (arguably) provide the highest level of insulation. However, these methods are reliant on mechanical parts, which do wear down over time.

Example Analog Isolation Products

While DIY solutions can provide engineers with a means to an end for the problem at hand, it is often more important for the designer to instead focus on the design as a whole. Therefore, it is often best for engineers to look at analog isolation solutions that are already on the market, as these products will save designers time and money.

The AMC1301 is an isolation amplifier that has a low offset error and drift, a fixed gain of 8.2, low non-linearity, and 3.3V operation on both low side and high-side. It uses an internal ADC to convert the read analog signal into a digital signal, which is then isolated across a barrier and then reformed into an analog signal with the use of a DAC. As the AMC1301 uses a digital method, the isolation amplifier is immune to magnetic fields and therefore applicable to environments where strong magnetic fields may exist.

The ADUM3190WSRQZ is an isolation amplifier that is similar to the AMC1301 in that it uses a receiver / transmitter pair for isolation. However, the ADUM3190WSRQZ uses a transformer (most likely a voltage-to-frequency method) as its method for isolation. This isolation amplifier has a 0.5% initial accuracy, can isolate up to 2.5kV, and offers a wide temperature range. However, the use of the internal transformer means that this amplifier is vulnerable to strong magnetic fields.

Conclusion: What is Isolation in Electronics?

Analog isolation is no easy feat and exactly how a designer chooses to accomplish this is entirely up to them. A bought solution is easy to implement but can be expensive depending on the solution used, while a custom solution can be time consuming and far more complex.

Most applications only require protection – as opposed to isolation – and so it is important that designers recognize when isolation is an absolute necessity.