ESD (Electrostatic Discharge) is one of the main enemies of electronic devices. The term is used to describe the phenomenon which appears between two objects at differently electrical potentials. Extensive measures have been taken in the industry to protect the devices being manufactured against the discharge (of which the most spectacular phenomenon is the spark itself). In any PCB or back-end assembly facility you will see people wearing cotton coats, heals and wrist straps to ground, or even special ESD flip-flops (or slippers). All these however, will not be able to ensure the ESD protection of the device once it is in the field. For that to happened, the electronics will have to be protected against high-value pulses at a design stage, and to employ methods and components that will ensure the device under question will survive the highest possible discharges that will potentially occur in the operating environment.

The purpose of this article is to give an overview of the ESD protection devices that a hardware engineer could consider for protecting the circuitry on the PCBs. The ESD problem is becoming more and more common, due to the reduction in size of components and of the transistors carved in silicon. Due to this fact, many technologies have been developed to offer protection. It is not always easy to select the proper one for an application, and the main criteria in considering which is best should be really based on the environment the application is intended to operate in; this will directly reflect in the requirements of the circuit to be protected.
At a PCB level, the ESD is translated by fast transients: that is high voltage signals with very steep slopes. By high voltage we mean several kilovolts that would be able to destroy the junctions of more sensitive silicon devices. This is rather due to the very high voltage value of the pulse, rather than due to its energy. This being the situation, the ESD protection devices are not really required to withstand high energy dissipation, but rather take the brunt of fast dv/dt slopes.

Below there is a short list (not complete) of considerations that need to be taken into account when designing for ESD protection:

a)If you use a voltage clamping device, the clamping voltage should always be smaller than whatever voltage the protected circuit can withstand on its own.
b)Depending on application, the voltage-claming device should be capable of handling several kV of a fast slope signal
c)It is desirable for the ESD protection device to be as small as possible to fit on the PCB as close as possible to the connector.
d)Always go for voltage ESD protection devices; the current ESD protection devices are generally not effective, as the current during the discharge is very small. It is the voltage that destroys components during ESD, not the current

The first 5 most common ESD protection methods and devices at a PCB level are:

1.The Capacitor
2.The Zener Diode
3.The Schottky Diode
4.The Transient Suppression Diode (also known as TSV)
5.The Varistor

I will try to briefly explain the situation in which each of these should be used.

The Capacitor

Symbol:

The type of capacitor that is used for ESD is the ceramic capacitor. It is preferred by many designers as an ESD protection method as it is simple and cheap. The drawback is the limited protection they offer, not being able to take on high spikes. Many of them get damaged to as low as 5kV transients (which is not that high in terms of discharge).
There is another disadvantage of the capacitors which is not so straightforward (like a burned cap). The energy discharged by the pulse is not transformed in heat, but it is taken by the capacitor and redirected from the protected device to the ground plane. This gives the current generated by the ESD voltage a low impedance path through where it can wreak havoc in unknown circumstances.

The Zener Diode 

Symbol: 

The initial purpose of these devices was to regulate voltage, rather than to protect against ESD. The main criterion driving their use for such a purpose, though, is their price (generally low). In terms of efficiency against surges the zener is better than a capacitor (ceramic). They are not as good from a clamping ratio point of view. The clamping ratio is defined as being the ratio between the clamping voltage and the breakdown voltage (DC), so finding a proper zener to clamp the voltage at the required levels of the protected circuit, while ensuring the diode itself will be safe, is a little more difficult.

Another drawback of the zener is its rather slow response to surge impulses, which in some circumstances might be critical. This is more obvious with pulses which are destructive due to their abrupt slopes rather than due to their voltage level.

The Schottky Diode

Symbol:

The Schottky diodes are used as ESD protections only for portions of the PCB. They see a general use for protecting the pins of the ICs against voltage pulses which are above the supply voltage or below the ground potential, and the way to employ them is in pairs, on the same data line:

Whenever a voltage higher then VCC will appear on the data line, the upper diode will be biased and will ensure that the voltage at the pin of the micro will be clamped at no more than VCC+0.5V (assuming a 0.5V voltage drop on the diode). Whenever a voltage drop lower than GND will appear on the data line, the lower diode will be forward biased and will ensure that the voltage at the pin of the micro will be clamped at no more than -0.5V.

Nowadays, the microcontrollers generally have these diodes integrated, but sometimes they are rated for a current much too small, and some external ones are needed; this method is not limited to microcontrollers: basically any pin of any IC can be offered some protection through this method.
It must be noted that very fast ESD events (in the range of nanoseconds) will be to fast for the schottky diodes to handle.

The Transient Suppression Diode (also known as TSV)

Symbol (for bidirectional TSV):

The TSV is also generally known as transzorb, or even transil (sometimes spelled as tranzil). Amongst their advantages, we can count the increased resistance to a surge and a decreased clamping ratio compared against the zener. A disadvantage of the TSV, like that of every other type of diode, is its slow response (fast slope ESD events will have a good chance of bypassing it).

For the TSVs a particular interest is represented by the peak pulse power the device can handle without breaking down. Even if the manufacturer of a TSV will generally provide a power rating for the device, this is not the general power that you would normally find dissipated on a component. In order to decide if the TSV will withstand the required pulses (for instance, in the automotive industry the potential destructive pulses are known in terms of shape and power) the datasheet must be consulted and this information extracted from the graphs provided by the manufacturer of the diode. For example, below are the peak pulse power vs pulse width and the peak pulse current vs time graphs of an automotive TSV:

Care should be taken in case one plans to employ a TSV in the same way like the schottky diodes on a data line (especially in case of high-speed data lines). The TSV generally presents a high parasitic capacitance which would adversely influence the slopes of the data signal, thus corrupting the information.

The Varistor

Symbol:

The name of the varistor comes from “variable resistor”, and this is due to the non-ohmic voltage-current characteristic of this device. It solves some of the problems generally presented by the diodes, as the varistor is known to be a fast-response device. Another advantage (but this is by no means exclusive to the varistor) is the small size of the package it may be found in (as small as 0402). These, together with the ruggedness against surges which is conferred to the varistor by the ceramic material used in manufacturing it, make it one of the best ESD protection methods, as they will withstand even the highest voltage spikes in terms of ESD.

The parasitic capacitance of the varistor is also lower, due to the possibility of adjusting the sizes of the electrodes; this makes them suitable also for ESD protection of high speed data lines.
One should be careful when employing varistors, as they are only meant to shunt high power only for very short time intervals (a few microseconds, at most); a lightning strike would generate such an ESD surge, for instance. The varistor is not, however, capable of conducting sustained energy, which is a case that might occur given some specific circumstances on the utility power grid (like loss of a neutral conductor, or shorted lines in a high voltage system).

As a general conclusion I would want to point out that the ESD protection methods discussed in this article are by no means a complete list. They are only PCB level solutions, suitable for protection during assembly, handling and operation in a controlled environment. None of the presented methods would be capable of withstanding a direct lightning strike or surges that last long (more than a few milliseconds, in a best case!).