Article
February 1, 2023
A Look Into Generator Waveforms: Do They Meet the IEC 61000-4-2 Waveform Specification?

This article explores the waveform specifications called out in the IEC 61000-4-2 standard [1]. Verifying that a generator meets this specification requires specific equipment and conversion equations. The setup, data collection, and calculations required to validate equipment are explained here. Waveforms have been captured using the target called out in the standard, often called a Pelligrini target, for three different generator manufacturers and one pulser. These waveforms are analyzed in the time domain to verify the generators were within the standard’s specification. What comes as a surprise is the real waveforms. Waveforms from different generators look very different and it is surprising that they all pass the specification [2].

IEC Specification

The IEC 61000-4-2 waveform is shown in Figure 1 for contact discharge mode. The parameters that are called out in the specification are rise time, peak current, 30 ns current, and 60 ns current.

Figure 1: Time Domain Specs for IEC Waveform

The spec limits are shown in Table 1. Note that the spec limits are quite wide; for instance, at the 30 ns and 60 ns current, the waveforms can vary by ±30%. The peak current can even vary by ±15%.

Pulse Parameter Value Unit
10 to 90% Rise Time 0.8 +/- 25% ns
First Peak Current (Ip) 3.75 +/- 15% A / kV
Current at 30 ns from initial 10% point 2 +/- 30% A / kV
Current at 60 ns from initial 10% point 1 +/- 30% A / kV

Table 1: IEC Pulse Parameters and Spec Limits

Test Set-Up

The test setup to capture the waveforms is shown in Figure 2. It uses a target for the tip of the ESD Generator to contact and thus capture the waveform. This target is seated in a large ground plane; thus, radiated fields will not be captured and are not a part of this study. A faraday cage can also be used to house the oscilloscope.

On the other side of the target is a high frequency cable connecting to ample amounts of attenuation to protect the front end of a high-speed oscilloscope. The attenuators in Figure 2 are 50-watt attenuators with a total attenuation of 29dB. This amount of attenuation ensures the peak voltage incident on the scope does not exceed the max voltage of 5V RMS on any precharge voltage setting. The oscilloscope should be at least 6 GHz of bandwidth in order to measure the rise time accurately. Adequate sampling is also needed to ensure enough points during the rise time for analysis. Typically, the scope is set to 20 nsec or 40 nsec per division. Adjustment of the vertical scale is needed to always capture the waveform using the entire vertical screen of the scope. This further ensures that the vertical resolution is at its maximum.

Figure 2: Test Set-up to Capture Generator Waveforms

Figure 3a shows a closeup view of an example target from the front and Figure 3b shows the back of the target with an N to SMA adapter for connection to a high frequency cable.

Figure 3: Example Target, a) Front b) Back

Data Collection and Current Waveform Calculation

To capture a waveform the scope is set to single shot mode with the trigger on a positive rising edge. The waveform is stored for further processing. The data that the scope stores is represented in Figure 4 as the term Vscope. This data has to be converted to current in the generator. Figure 4 shows the circuit representation of the test set-up in Figure 2. The term Vcorr is the voltage incident on the attenuator before the scope input. The equation in Figure 4 shows how to convert the scope data to this voltage Vcorr.

Figure 4: Detail of Circuit for Capturing Waveforms

Figure 5 shows the equivalent circuit of Figure 4 with the input to the attenuator as 50 ohms to ground. Below this equivalent circuit are the calculations to transform the corrected voltage from Figure 5 to pulse current. The target used in this study had 2 ohms to ground and a series 48 ohms connecting to output on the back of the target. Icorr, is effectively the current into the attenuator. This current can then be used to calculate the voltage the generator produces across the 2 ohms to ground which is Vpulse in the figure. This voltage is divided by the target impedance yielding the target current. The target current plus the current into the attenuator is the current in the generator. This is the current called out in the IEC specification. The final equation in Figure 5 shows conversion from the scope data, Vscope, to the generator current, Ipulse.

Figure 5: Equations to Calculate Current in the Target

Real Current Waveforms

Although Figure 1 looks well behaved, actual waveforms are not. Figure 6a shows three different generator waveforms at 8kV and even a waveform from a 50-ohm pulser. These waveforms were captured using a scope with a bandwidth of 7GHz and a sampling rate of 20 Gsamples/sec. The scope was set up so that the max record length was used while capturing 400 ns worth of data. All of these generators meet the IEC spec. Figure 6b shows that although all the waveforms comply with the rise time, peak current, 30 ns, and 60 ns current, they still vary considerably and can look quite different from each other [3].

Figure 6: Current waveforms for 3 IEC Generators.

Summary

Verifying that a generator meets the IEC spec requires a target, attenuation, and a high-speed scope. Using the scope data to calculate the current a generator has in its waveform has been explained. Despite the smooth waveform shown in the IEC standard, real generator waveforms are very erratic but nonetheless compliant. Pulsers, on the other hand, do provide an extremely repeatable and smooth waveform.

References

  1. “IEC 61000-4-2: Electromagnetic Compatibility (EMC) – Part 4-2: Testing and Measurement Techniques – Electrostatic Discharge Immunity Test,” IEC International Standard, 2007.
  2. K. Muhonen et. al, “HMM Round Robin Study: What to Expect When Testing Components to the IEC 6100-4-2 Standard Waveform,” EOS/ESD Symposium, 2012.
  3. K. Muhonen, et al., “Spectral Analysis of IEC Generators for System Level Testing of RF Components,” EOS/ESD Symposium, 2020.