Enlarge this picture Fig. 1. ESD conducted path.

Wintertime is prime time for electrostatic discharge (ESD) activity, at least in the Northern states of the U.S. Colder temperatures mean lower humidity, resulting in an increased number of ESD events. These incidents are especially troublesome to modern electronics, causing functional upset and damage. There are a number of problems associated with ESD, but, fortunately, also some solutions.

ESD as an EMI Problem. Before exploring specifics, it’s important to note that, since even small amounts of ESD can damage modern electronic components, most electronics manufacturers have extensive ESD programs in place. These include ionizers, conductive materials (floors, tables, etc.) and wrist straps. The goal in manufacturing is to prevent even a single discharge.

In the EMI world, this prevention of a discharge is not considered an EMI problem. When electronic equipment is placed in use, however, one can no longer guarantee that ESD events won’t occur. And when such a discharge occurs, with rise times of 1 nSec or less, it can, and often does, become a major EMI problem.

ESD as an EMI Source. Most ESD events result from tribolectric charging. This is commonly created by the separation of two dissimilar materials, but it also occurs with similar materials, as with paper unrolling. Examples of charge build-up include a walking person (shoes contact and then separate from the floor surface), a rolling cart (wheel contacts and separates from the floor), or even a moving integrated circuit (part contacts and separates from a plastic container).

The primary concern is human-generated ESD, because humans most often touch and use equipment incorporating electronic components, which today includes most modern appliances. It may take seconds or minutes to charge up a human, but the discharge lasts perhaps 20 nS, with currents of 10 A or more. There is not enough energy in a discharge to harm a human, but more than enough to upset or damage electronics.

The leading edge of a human discharge has a 1 nS rise time, which is equivalent to 300 MHz of interference, which is very fast, indeed.

ESD and EMI Coupling Paths. ESD can attack electronics via two major paths — directly by conduction and indirectly by electromagnetic radiation. Fig. 1 shows the conducted path. Direct injection into a signal pin (input or output) is destructive, making circuit protection mandatory.

Injection into the ground path is more common. Currents encountering even small ground impedances result in ground bounce, generally causing signal errors. Fig. 2 shows an electric field and a magnetic field coupling path, both common in ESD problems. In fact, the electric field radiation path is now part of the commercial ESD test procedures, and is referred to as the “indirect” test method. By the way, this field coupling is very real. It has been observed that failures due to the transient fields have occurred up to 20 feet away.

These multiple paths often mean that the design engineer may need multiple sets of design fixes: filtering or transient protection for the input/output pins; ferrites or other current limiting for the power/ground paths; and shielding for the radiated path.

For equipment with cables, such as computers, component TV and sound systems, ESD effects upon input/output cables represent a major issue, largely because the wire is attached directly to sensitive electronics. For standalone equipment, such as most major appliances, there will not be any external cables except for the AC power. In those cases, the user interface or operator controls are the major issue. For all equipment, direct discharge to metallic surfaces or protrusions will be a problem. ESD and EMI Victims. As already mentioned, ESD can result in both damage and upset. Damage is likely to occur when ESD is injected directly on any unprotected I/O pin of any electronic device, digital or analog. Upset is more likely to affect only fast digital circuits, since most analog circuits are too slow to respond to ESD events.

Circuits that are particularly vulnerable to ESD include resets, interrupts and control circuits. Unwanted resets are a very common ESD problem, so common that users should routinely add components to reset circuits to prevent ESD induced events. Interrupts, particularly non-maskable interrupts, often warrant the same attention. Control lines, such as memory read/write, are less often a problem, but may also require protection. One peculiarity in PROMs is that memory may be corrupted, even if the write hardware is not on-board.

Even power circuits may also be vulnerable to ESD. We recently chased an ESD problem that repeatedly caused a power supply to shut down due to false triggering on an input overvoltage protection circuit.

ESD Design Solutions. When analyzing ESD events, it is more useful to consider current rather than voltage. An ESD event is like a dam breaking in the mountains. After the dam breaks, the pressure (voltage) may be low, but it is the gushing water (current) that causes the damage.

If the current encounters a high impedance, a second high voltage may result. If it is too high, a voltage “punch through” can occur. If the current encounters low impedances, the high currents can cause damage by simple heating.

Given those circumstances, the designer has three choices when dealing with ESD: prevention, diversion and limiting.

Prevention

The most effective way to handle ESD is to prevent discharge from occurring by using insulation or spatial separation. One should keep in mind that at 15 kV, an ESD arc can easily jump over one centimeter in air, and even farther if sharp points (such as screws) are present.

Human ESD most often results from contact with the equipment, most commonly with operator controls. ESD won’t penetrate insulators if sufficiently thick, including most plastics or even a thick coat of paint. But it will find air paths, no matter how minute and circuitous. To prevent discharge, it is necessary to recess metallic parts. It is also necessary to verify that the dielectric strength of the controls is sufficient to prevent discharge. While most plastics do fairly well, some plastics will not withstand even modest ESD hits. Once a path has been burned through any dielectric, a low breakdown path exists, increasing the vulnerability for the same location.

While a thick coat of paint, such as found with modern clothes washers, will block a discharge, it takes only a small pinhole or nick in the paint to establish a path. That’s important to note because there is often some bare metal where the fasteners are placed.

ESD will travel surprising distances along a material path, due to surface conditions.

Diversion

Enlarge this picture Fig. 2. ESD coupling paths.

If discharge cannot be prevented, the next best option is to divert ESD current away from a critical circuit input or output. Operator controls are a particular problem. Discharge currents that get to the control will follow the wire to whatever that control is connected to unless diverted in some manner. The best solution is to ground the switches to the enclosure to divert the currents. If the switches cannot be grounded, they can be diverted by a transient protector or small capacitor (typically 1,000 pF to 10,000 pF).

Transient protectors must be fast enough to shunt ESD currents, such as silicon protectors (Zener diodes) or surface mount MOVs (metal oxide varistors). Regular MOVs such as used for power surges and arc-gap devices are generally too slow to shunt ESD currents. Capacitors need to be mounted so as to minimize series inductance. Lead lengths of even a half inch are too much.

RFI shielding is the best method of diverting ESD currents. Done well, ESD currents and fields are blocked from penetrating the enclosure. Any metal will work well, but the gaps in the opening should be kept to a minimum. Openings greater than a few inches can be a problem, and mating surface must be conductive. Basically, hanging sheet metal on a frame does not constitute shielding. The most intense fields occur close to the gaps in the shield, so avoid placing cables near the gaps.

Limiting

This is often used to prevent dumping large currents into the circuit board power or ground distribution. On I/O lines, having 100 ohms of impedance (using a ferrite or resistor) facing the arc source is adequate to limit ESD currents through a shunt capacitor or transient protector. Separate ferrites may be needed on power or ground lines if they are exposed directly to ESD discharges.

Testing

It is not always necessary to go to an EMC lab or wait until the end of a project to begin performing useful ESD testing. As many “pre-compliance” tests should be run as possible during the design stage. The earlier one starts to uncover problems, the more time (and design flexibility) one has to fix them.

One can rent or buy an ESD “gun” that meets the appropriate ESD test recommendations. The procedures described in EN 61000-4-2, the European test method for ESD, are recommended for testing. Begin with the “indirect” tests, as they are less likely to cause damage. Once the product has been hardened for the indirect tests, then one can proceed to the direct tests.

In both cases, start with a low voltage and increase in 2 Kv increments. For ESD, it is not adequate to test only at the highest levels. There have been many cases of “windows of susceptibility” at lower voltage levels. Finally, remember that the direct tests may cause damage, so have spare parts available as needed.

Summary

Electrostatic discharge is a major threat to electronic equipment, causing functional upset and permanent damage. The best defense is to prevent discharge from occurring. If discharge cannot be avoided, then the currents need to be blocked from reaching vulnerable circuits, either by diverting the ESD currents or by blocking them.