Medical Device Link.

A common myth persists that once a device is mounted onto a printed circuit board (PCB), it is significantly less vulnerable to electrostatic discharge (ESD).1–3 However, it has long been known that integrated circuits (ICs) and other ESD-sensitive components remain at risk during board assembly and subsequent handling and installation.2–4 Nonetheless, the myth has been compounded by the fact that most ESD testing and characterization of these components has been done on stand-alone parts. Further, IC failure analysis data have caused many to conclude that ESD failures are relatively rare when compared with the number of other electrical failures commonly classified as electrical overstress (EOS). IC failure analysis data are based on knowledge of failure signatures seen in standard human body model and charged-device model (CDM) tests.

Recent data and experience reported by several companies and laboratories now indicate that many failures previously classified as EOS are instead the result of ESD failures due to charged-board events (CBEs) or cable-discharge events (CDEs).1 As their names imply, a CBE discharge occurs when a charged circuit board is suddenly discharged. A CDE occurs when a charged cable is suddenly discharged. Although this article focuses on CBE ESD failures, some of the analysis is also applicable to CDE ESD failures; each type of discharge can result in very-high-energy events that severely damage ICs and other components.

A charged board stores much more energy than a device (e.g., an IC) because its capacitance is many times larger. In fact, the charge (energy) transferred in the event is so large that it can cause EOS-like failures to the components on the board. Figure 1 illustrates considerably more-extensive damage resulting from the CBE than CDM of the component separately. The vast majority of failure analysis experts would incorrectly diagnose the CBE damage as EOS.

Charged-Board Event: Field Returns

The following investigation provides an example of ICs that were robust to ESD at the component level but were nonetheless damaged by ESD at the board level. The damage was simulated via field-induced charged-board event (FICBE) testing using a conventional CDM test system. For a given charge voltage, the CBE discharge waveforms had much higher peak currents than the corresponding CDM discharge waveforms. Therefore, the CBE damage was more severe than the CDM damage. In some cases, the CBE damage was so severe that it could easily be mistaken for EOS damage. The susceptibility of a given IC to CBE damage is a complex function of variables including the IC-on-chip protection network; the IC package design; the size of the power planes on the PCB; and the number of power supply pins on the IC tied to the power planes.

During system-level production testing and field application, there was an instance of a several hundred parts per million failure rate on a four-level-metal, deep-submicron CMOS digital signal processor (DSP) packaged in a 28 × 28-mm, 208-lead plastic quad flat pack. The failure modes varied, but typically involved functional failures within a small block of circuitry within the DSP.

The DSP was located near a corner of the PCB above relatively large copper ground planes in both the top and bottom layers of the PCB. All the components were on one side of a four-layer PCB. The top and bottom ground planes were interconnected by numerous plated through-holes. The 35 GND pins on the DSP package were tied to the interconnected copper ground planes, while the 33 Vdd pins were tied to an internal copper Vdd power plane. The other internal PCB layer was used for routing I/O signals.

FICBE testing was conducted using a KeyTek Verifier robotic CDM test system. To assist simulating the failures, a customer provided numerous PCBs. Since the complete PCB was larger than the 127-mm (5-in.) field charging plate, the PCBs were cut down in size. However, the ground plane under the DSP was kept fully intact. Because full electrical testing of the cut-down PCBs was not feasible, the DSP on each PCB was decapsulated to expose the die for visual inspection purposes. Initial high-magnification die inspection of the DSPs on the PCBs revealed no anomalies like those previously noted.

The steps for the FICBE test method were as follows:

1. The cut-down PCB was centered on the charging plate (see Figure 2). Fortunately, the PCB had no components on the bottom side, so it rested flat on the charging plate. In this configuration, the capacitance measured between the PCB ground planes and the charging plate was ~420 pF, while the capacitance between the PCB Vdd plane and the charging plate was ~460 pF.

2. The charging plate was raised to 125 V, and then the ground plane was discharged at a test pad close to the edge of the PCB. Consistent with the methodology in ESD Association STM5.3.1-1999, this step was repeated two more times.5

3. High magnification optical die inspection was conducted to look for the onset of damage.

4. The charging plate was brought to –125 V and then the ground plane was discharged at the same PCB location. This was repeated two more times.

5. High-magnification die inspection was again conducted to look for the onset of damage.

6. Consistent with the procedure in Steps 2–5, the ground plane on the same PCB was subjected to FICBE testing in 125-V charge voltage increments until high-magnification optical inspection revealed damage.

DSP IC Failure Analysis

The damage resulting from a 1-kV stressing voltage is shown in Figure 3. The resulting discharge waveforms are shown in Figure 4. The CBE discharge current is approximately 9 A while the CDM peak current of the device alone is under 4 A. These stresses frequently exceed the design capabilities of ICs. The damage is far more severe than what is typically observed for CDM device stress testing (JEDEC C101); it would easily be misdiagnosed by failure analysis experts.6

Based on 10 years of data at the author’s company, it appears that approximately 50% of supposed real-world EOS failures are actually ESD at the circuit board level (CBE) or at the system level (CDE). Such a realization is especially significant because so many failure analysis reports indict EOS as the root cause when, more likely, ESD is the true failure mechanism. These incorrect diagnostics trigger costly and ineffective corrective actions in manufacturing and in chip design. They also trigger dead-end investigations into test equipment and power supplies.

Charged-Board Event: Manufacturing

A separate experience illustrates how the CBE can take place in a manufacturing operation. Many printed wiring assemblies (circuit boards) include a cover or faceplate that provides a protective covering when the board is installed in a system (see Figure 5). To keep material costs down, they are often made of insulating plastic.

In this case, the board was in production for more than a year without any indication of a significant problem. Suddenly, however, the removal rate of a certain linear CMOS part began to rise. The failure rate averaged 2.5% with rates as high as 40% on certain days. The device failures were observed after circuit board testing, and the observed electrical signature was excessive leakage current between two pins on the device. The leakage was high enough to cause the circuit board to fail its functional requirements.

These observations and subsequent failure modes and effects analysis (FMEA) pointed strongly to CBE ESD as the source of the problem. The production line had a well-designed ESD program known to be in compliance with the current best practices. Further, a careful analysis of the line produced no indication of why this particular part was failing at higher than normal levels. Most significantly, the failing component was well designed with a CDM withstand voltage of 1500 V.

While investigating changes in the design and materials, it was learned that the source of the plastic for the faceplate had changed around the time that the failure levels began to increase. Both the base resin and the molder had changed. It was then found that the electrostatic voltages on the faceplates were extremely high, with 10 kV being typical, and that these voltages persisted for days or weeks. Laboratory investigations then showed that the faceplates from the new source tended to charge to levels about five times higher than the previous ones and that the charge retention was much longer as well.

For reasons that will become clear in this discussion, the board (device)-level failure mode analysis was difficult. Eventually, it was demonstrated that the exact failure that was observed in the factory could be produced by tribocharging the faceplate and then touching (grounding) the circuit board in a particular way. This was a classic example of a field-induced CBE (CDM) failure. Initial investigations into the failure mechanism of the circuit boards indicate that the pin (21 or 22) that failed in the factory was never physically touched during testing or handling of the circuit board. This was surprising because the CBE/CDM failure of a pin requires that the pin be grounded. Thus, further studies were conducted in the laboratory, which confirmed that the pin 21-22 leakage current could be produced by touching a pin on a transformer mounted near the faceplate.

Pins 21 and 22 were connected by a low-resistance bus on the PCB to pin 36 of the CMOS device (Figure 5). Thus, the pin that exhibited the failure was different than the pin stressed. This is not unusual for CDM events. However, this is seldom observed in routine qualification of devices because poststress testing of the device is usually only done after all pins have been stressed. Therefore, it is important in FMEA investigations to stress the device in a manner that resembles as closely as possible the actual sequence of real-world events; such a procedure is essential to confirm the failure mechanism.

The next step was to understand how the charging and discharging events were occurring in the factory. With the aid of ESD event detectors, the discharges were found to be occurring during the testing of the circuit board. The entire scenario is represented schematically in Figure 5. When the circuit board with its charged faceplate was placed in the tester, the first test probe to touch the board touched a pin on the transformer near the charged faceplate. The transformer pin was approximately a half inch from the charged faceplate. Because the voltages on the faceplates were very high, the effective induced voltage as seen by the board resulted in a discharge current that exceeded the device capability. As noted above, these discharge currents often far exceed the design capability of either the components or circuit boards. Thus, sound CBE manufacturing practices are both essential and the baseline of defense. It is no longer possible to rely just on designed-in protection.

Conclusion

Misdiagnosis of EOS failures is a widespread and costly problem within the electronics industry. Where careful data have been collected, it has been demonstrated that approximately 50% of the so-called EOS failures are actually ESD at the board level (CBE) or system level (CDE).

These data suggest that failure analysts should give stronger consideration to these types of board-level and system-level ESD events before assigning an EOS diagnosis to the failure. This will support more-effective root-cause analysis and prevention of these failures. The discharge currents in many cases exceed the design capabilities of either ICs or circuit boards. Thus, manufacturing and field ESD control methods become critically important and are the main lines of defense.

Because the CBE is nearly the same mechanism as CDM, the factory and field control methods are similar to those for CDM. The diagnostics involve EMI ESD event detection and warrant a full understanding of the physics of CDM and ESD by induction.

Ted Dangelmayer is president and CEO and Terry Welsher is senior vice president at Dangelmayer Associates LLC (Gloucester, MA). They can be contacted at ted@dangelmayer.com and terry@dangelmayer.com, respectively. Andrew Olney is director of reliability, product analysis, calibration, and ESD at Analog Devices Inc. (Wilmington, MA). He can be reached at andrew.com. ■