Jeremy M. Smallwood - The ESD Control Program Handbook

PCBs often enter a production process highly charged. They can remain charged for long periods or can become charged as they are transported or go through a handing or assembly process. Voltages on the board up to 1000 V are not unusual, although the measured voltage will typically change with the proximity of the PCB to other objects. A PCB can also have high induced voltage if there is a highly charged insulator or another source of electrostatic field nearby.

If a conductor (e.g. track or component pin) on the charged board touches a highly conductive machine part (e.g. stop pin), a charged board ESD event can occur ( Figure 2.17). The PCB can have high effective capacitance, so this type of discharge can be quite energetic.

Many products, modules, or subassemblies have an insulating plastic housing containing a circuit board. The connections to this may be brought out to terminations at flying leads or a connector.

The housing can become highly charged, e.g. by rubbing or removal from packaging, inducing a high voltage on the PCB within the housing ( Figure 2.18). If a connection is made to the module in this state, a discharge can occur at the termination at which contact is made.

Figure 2.17ESD waveform from a printed circuit board (above) charged to 1 kV (below) field induced charged by insulator 40 mm away.

Figure 2.18ESD waveform from a charged automotive module taken out of a polythene bag. Charge transferred 35 nC.

Figure 2.19Voltage on an automotive cable core as polythene packaging is removed.

Cables and wiring looms can have significant capacitance between the wires in the cable and between the wires and ground. This can be of the order of 100 pFm −1. Wires in the cable can become charged by various means such as by movement of the cable or by removal of the cable from packaging ( Figure 2.19). If the cable is connected to equipment in this state, a charged cable ESD event can occur to the first terminal to make a connection ( Figure 2.20).

Many ESD sources can be simply modeled using a simple R‐L‐C circuit ( Figure 2.21). The values of each component vary widely between different sources and help to explain the different types of waveforms observed.

At the heart of any ESD source is charge build‐up and storage. This is represented by the capacitance in the model C . In many cases in real life, this charge storage may be on a conductor (e.g. metal item).

Figure 2.20ESD waveform from a charged automotive wiring loom cable lying against an earthed metal plate. Positive (above) and negative (below) charging polarity.

Figure 2.21Electronic model of a simple ESD source.

The discharge is usually initiated by a breakdown of an air gap or some other insulating medium. At low voltages, it can also be initiated by contact or near‐contact between two conductors. The discharge can itself have significant impedance R ESDthat can affect the waveforms produced and the energy delivered into the victim device. Often, however, this is negligible compared to the other impedances in the circuit, especially for larger ESD events.

After the discharge commences, the current flows through some circuit that includes some elements of resistance R sand inductance L s. These are normally due to the resistance and electrical properties of the materials in the current path.

In the case of ESD to a victim device, the device also has impedance, modeled in this simple circuit by a resistance R d. In practice, a nonlinear impedance would be more typical of a semiconductor device. The impedance of the spark channel is highly variable and nonlinear.

For simplification, the total circuit resistance R is assumed to be linear and is the sum of the circuit resistances.

The discharge current I ESDof this circuit has the form

For derivation of the equations for this and the following equations, the reader is referred to other texts (e.g. Agarwal and Lang 1987, https://en.wikipedia.org/wiki/RLC_circuit). This equation has two roots α, β given by

The waveform shape takes very different forms depending on the circuit component values. If the total circuit resistance is large and dominates the discharge path impedance, the waveform has a unidirectional shape, simulated in Figure 2.22using model component values given for human‐body model ESD (see Table 3.12). This occurs when

Figure 2.22Simulated overdamped device current waveform IESD for dominant circuit resistance: R s= 1500 Ω, R d= 10 Ω, R ESD= 0 Ω, L s= 10 000 nH, C s= 100 pF, V ESD= 500 V.

The discharge current rises rapidly to a peak I pthat, when inductance is small, approaches the value and polarity near that predicted by Ohms law.

Thereafter, the current drops nearly exponentially with decay time approaching R s C ESD.

At the other extreme, if the circuit resistance is insignificant compared to the inductive and capacitive impedance, the waveform is quite different. This occurs when

The waveform rises to a peak and then oscillates negative and positive about zero. The overall amplitude decays exponentially with time, simulated in Figure 2.23using values given for machine model ESD (Table 3.12).

Between the two extremes, the waveform duration decreases and is minimum around the point of critical damped waveform, where the waveform changes between the two different shape types. This is simulated in Figure 2.24using model values close to those given for the charged device model (Table 3.12). This occurs at the condition