Coupling-Decoupling Networks (CDNs) are the unglamorous workhorses that get “hitched up” to an impulse generator in order to perform impulse testing on powered equipment. As such, performance and calibration requirements for CDNs have not been specified to the same level of detail as the impulse generator. This is no longer the case with the Third Edition of IEC 61000-4-5, the international standard for surge immunity testing.

What's New (and What's Not)

Figure 1: Differential-mode coupling on power lines using 18 uF capacitor (decoupling circuit also shown).

The Third Edition specifies new calibration requirements for the CDN: section 6.4 for CDNs used with 1.2×50/8×20 uS impulse generators, and section A3 in Annex A for CDNs used with 10×700/5×320 uS generators. The first type of generator produces an impulse with a nominal 1.2 uS rise time, and a 50 uS duration into an open circuit (high impedance) load. This same generator produces an 8 uS rise, 20 uS duration current waveform into a short circuit. This generator is used for all impulse tests in IEC 61000-4-5 except for tests on outdoor symmetrical communication lines.

The second type of generator (for outdoor symmetrical communication lines) produces an impulse that has a 10 uS rise time and 700 uS duration into an open circuit, and a 5 uS rise, 320 uS duration short-circuit current. This longer-wavelength impulse is considered more representative of the type of waveform that would be capable of traveling long distances. For the CDN, the components used, and the method of injecting the impulse onto the signal or power lines, differ for the two types of generators.

Figure 2: Common-mode coupling on power lines using 9 uF + 10 Ohms (decoupling circuit also shown).

The basic components for power-line coupling and decoupling of the 1.2×50/8×20 waveform are unchanged between the Second and Third Edition of IEC 61000-4-5. For differential-mode testing (from Line to Line, or Line to Neutral), an 18 uF capacitor is used as the coupling element (Figure 1). For common-mode testing (Line to Ground) a 9 uF capacitor in series with a 10-Ohm resistor is used (Figure 2). The impact of the Third Edition changes is felt primarily in the decoupling elements of the CDN, which are also shown in Figures 1 and 2.

The goal of the decoupling network is twofold. First, allow AC or DC power to be delivered from the AE (Auxiliary Equipment) port to the EUT (Equipment Under Test) port, so the decoupling network must be a low impedance to DC and low-frequency AC. The second goal is for the AE port (as seen through the decoupling components) to appear as a high impedance to the high-frequency impulse. This can be expressed in simple mathematical terms, which generally apply for all decoupling circuits:

For power and signals: (decoupling circuit impedance) << (EUT impedance)

For the impulse waveform: (decoupling circuit impedance) >> (EUT impedance)

CDN Calibration

The Third Edition adds new constraints to the decoupling circuit in sections 6.4.2 (power lines) and 6.4.3 (interconnection lines, also known as data or signal lines). Waveform characteristics are specified for the EUT port and must be met with the CDN connected to the impulse generator. For AC/DC power ports, there are additional requirements for the residual impulse that are presented at the AE port (residual impulse on the AE port must be relatively small). These two requirements are simply a quantitative statement (with limits specified) of what was previously stated: To the impulse waveform, the decoupling circuit impedance must be much greater than the EUT impedance. If this is the case, then most of the impulse voltage and energy will be applied to the EUT, and very little impulse voltage/energy will travel back to the AE port.

Table A.3 from IEC 61000-4-5 Third Edition

A limit is placed on the size of the inductors used in the decoupling network: 1.5 mH for power lines (the inductors shown in Figures 1 and 2). On the other hand, the capacitors in the decoupling circuit may seem relatively benign, as there is no specification in the standard for these components. However, there are conflicting constraints on these capacitors that are imposed by the requirements in the standard: prevent excessive voltage from appearing at the AE port (capacitors must be LARGE), and at the same time, the voltage and energy must be delivered to the EUT port and not absorbed by the decoupling network (capacitors must be SMALL). The residual voltage at the AE port during an impulse test (with nothing connected to the AE port) is generally limited to 15% of the overall surge voltage. An excerpt from section 6.4.2 follows:

“The residual surge voltage measured between surged lines and ground on the a.c./d.c. power port of the decoupling network with EUT and mains supply not connected shall not exceed 15% of the maximum applied test voltage or twice the rated peak voltage of the CDN, whichever is higher... All performance characteristics... shall be met at the output of the CDN with the a.c./d.c. power port open-circuit.”

During an actual test (with equipment connected to the EUT port, and power applied to the AE port), the AE-connected equipment will likely appear as a lower impedance at the AE port of the CDN. The effect of this circuit change will have a limited impact on the impulse voltage/energy that is delivered to the EUT, because the CDN itself must primarily block the impulse from traveling to the AE port, as opposed to clamping the impulse.

Similar constraints are placed on the decoupling networks used with unsymmetrical and symmetrical interconnection lines. Generally, capacitors are not used at all in this configuration, which can result in unwanted/undesirable impulse voltage and energy being presented at the AE port. The Third Edition standard warns the reader of this situation in section

“The residual surge voltage measured between the surged lines and ground on the AE side of the CDN, with the EUT and AE equipment disconnected, shall be measured and recorded so that users of the CDN may determine if the protection is sufficient for use with a particular AE.”

The calibration tests specified in section 6.4.2 of the Third Edition for power ports are performed with the AE ports open-circuit. Conversely, the calibration tests performed on interconnect lines are run with all AE lines short-circuited and connected to ground. The concern for surges conducted on interconnect lines is that the decoupling network should not absorb voltage/energy and thus reduce the voltage/energy that is applied to the EUT. This limits the type and size of protective components used in the CDN.

The requirements for CDNs used with the 10×700/5×320 impulse generator are in section A.3 of the Third Edition. Again, the concern that has been addressed in the new requirements is that the CDN should not reduce the level of voltage/energy that is delivered to the EUT. Table A.3 in the standard outlines the calibration steps, which are similar to the steps for calibrating the 1.2×50/8×20 CDN. Emphasis in the standard is placed on preventing the CDN from reducing the voltage/energy delivered to the EUT. This is a practical concern if an inappropriate decoupling network is used during impulse testing. For example, if the decoupling inductors were to saturate during an impulse test, they would change from a high impedance (blocking the impulse) to a low impedance, allowing the surge voltage/energy to travel to the AE port instead of the EUT port. The effect would be twofold: the EUT is subjected to a lower-than-expected surge, thus allowing a product to more easily pass the impulse test, and the AE port is subjected to impulse energy that could potentially damage connected equipment.


The Third Edition of IEC 61000-4-5 should allow for more consistent and repeatable impulse testing when CDNs are used. As these new Third Edition requirements are put into practice, expect to see CDNs that do less to protect the AE port-connected equipment from damage in order to meet the requirement to minimize the reduction in voltage/energy to the EUT. With the exception of AC/DC power port testing, the burden of protecting equipment connected to the AE port is left to the end user.

This article was written by Jeff Gray, Chief Technology Officer at Compliance West USA, San Diego, CA. For more information, Click Here .