Medical Electronics Manufacturing Fall 1999

ELECTROMAGNETIC COMPATIBILITY

Understanding the issues surrounding IEC 61000-4-3 can ensure that the laboratory results are sufficient for the product being tested.

Clark Vitek

Since 1996, many medical electronic products have been required to meet standards for immunity to radiated electromagnetic fields. In Europe, radiated electromagnetic field immunity is a requirement for competent body approval or self-certification under the Medical Devices Directive. In most cases, the required test standard is IEC 61000-4-3, which is published in Europe as European Norm (EN) 61000-4-3. Although the test is not mandatory in the United States, FDA endorses it for respiratory and anesthesiology equipment, patient monitoring devices, and other medical electronics equipment.

A free-space (fully anechoic) test chamber. Photo courtesy of CKC Laboratories (Redmond, WA) and Medtronic Physio Control (Redmond, WA).

The current version of IEC 61000-4-3 is the second revision of the earlier IEC 801-3 (1984) test standard. This revision was the result of many years of harmonization and attempts to improve test repeatability. Publication of the revised standard as a European norm in 1995 expanded its applicability to almost every type of electronic equipment, including medical electronics. Unfortunately, even with the revisions, this test continues to present problems with test repeatability, interlaboratory correlation, and interpretation of the requirements and results. The International Electrotechnical Commission (IEC) is working on another revision that is designed to further clarify the areas that cause confusion. In the meantime, understanding some of the underlying problems inherent in conducting this test can help manufacturers better interpret the results provided by test laboratories.

Understanding these issues is important because a company's decision to declare that a product complies with IEC 61000-4-3 often depends on confidence in the results provided by an outside laboratory. This article provides a guide to some of the key issues surrounding the IEC 61000-4-3 test that can significantly affect the test results reported by laboratories. This guide should also help to evaluate a laboratory's competence in conducting and interpreting the tests for immunity to radiated electromagnetic fields.

Calibration of the Uniform Plane

IEC 61000-4-3 requires test facilities to provide a precalibrated area of 1.5 x 1.5 m. Further, the standard requires that 75% of the area must be within +6/–0 dB of the specified test level. This is often referred to as the field uniformity requirement. The uniform plane is a demonstration that a facility can provide the required field strength over a large area without distortions due to reflections in the room.

A free-space (fully anechoic) facility provides a more uniform field than a semianechoic facility. The difference is readily observable: a free-space facility has radio frequency (RF) absorber on all surfaces, including the floor, whereas a semianechoic facility has absorber on all surfaces except the floor. Because of the reflections from the floor, semianechoic facilities in general can distort the fields presented by up to 15 dB, or a factor of 5 times the intended field strength. This means that at some locations a test to the 10-V/m level might actually present more than 50 V/m to the equipment under test (EUT). The requirement to meet the 6-dB uniformity limit holds the maximum field distortion to a factor of 2 over 75% of the 1.5 x 1.5–m plane.

One might assume that all facilities that claim they meet the IEC 61000-4-3 field uniformity requirement are equally qualified to perform the test. Unfortunately, this is not the case. Note that the standard places a requirement on only 75% of the 1.5 x 1.5–m plane, so the variation allowable in the remaining 25% of the calibration area is completely unspecified. By controlling the reflections from the floor, a free-space facility is better designed to provide the required field uniformity over 100% of the uniform plane.

Another problem with uniform plane calibration is that the 1.5 x 1.5–m area is very large, and, in fact, for smaller equipment it is often more appropriate to calibrate the field over a much smaller area, such as 0.5 x 0.5 m. This is permitted in the standard, providing that a facility can meet the required +6/–0 dB uniformity over 100% of the calibration area. By using a smaller calibration area, the field generated will be closer to the specified test level (i.e., 3 V/m or 10 V/m) because it is easier to generate a uniform field over a smaller area, and because 100% of the data points must comply with the field uniformity requirement when a smaller area is used.

Compensation for Measurement Uncertainty

Many laboratories now are required to address measurement uncertainty to maintain accreditation in accordance with ISO Guide 25. Laboratories that are not accredited, however, most likely do not address the uncertainty question. Measurement uncertainty is never discussed or described in IEC 61000-4-3. However, compensating for measurement uncertainty when conducting the IEC 61000-4-3 test is not a trivial problem. In some cases, uncertainty can completely dominate the test results.

Field Uniformity. The first uncertainty issue arises during the uniform plane calibration. This calibration is performed using an isotropic field probe and a power meter to determine the system forward power required to generate the specified test level, such as 10 V/m. A typical field probe, operated properly at about 30–50% of full scale, still has a NIST-traceable uncertainty of about ±26%, or 2 dB.

When combined and expanded for other instrumentation used (such as a power meter and directional coupler), the total measurement uncertainty of the uniform plane calibration is typically around ±4 dB for a 95% statistical confidence. This means that if a laboratory is properly calibrated and delivering a 100% uniform field over the entire 1.5 x 1.5–m plane, the results are still subject to ±4 dB uncertainty traceable to NIST. This represents an uncertainty of ±58% in field strength delivered during the test before considering any nonuniformity in the 1.5 x 1.5–m plane.

Measurement Uncertainty. Laboratories are not currently required to compensate for measurement uncertainty when performing the IEC 61000-4-3 immunity test. To understand the potential effect of this, compare the 3-V/m test performed at two different facilities. To distinguish this problem from the field-uniformity issue discussed above (which is actually also a type of uncertainty), assume for these examples that both facilities provide perfect field uniformity over 100% of the area occupied by the EUT. During calibration of the uniform plane, Facility A provides the 3-V/m minimum level by calibrating to a nominal 4.74 V/m to compensate for the NIST-traceable calibration measurement uncertainty of ±4 dB (58%) above the required 3-V/m level. Because the uncertainty is ±4 dB (58%), the actual field strength delivered during a subsequent test will be between 3 V/m and 7.5 V/m traceable to NIST, meeting the required minimum field strength of 3 V/m. Now, take the case of Facility B, which might not consider measurement uncertainty during the calibration (remember, it is not required to). Facility B's calibration of a 3-V/m nominal field strength will represent an actual NIST-traceable field strength of between 1.875 V/m and 4.74 V/m during subsequent testing. The latter approach, therefore, might not satisfy the requirement to meet a 3-V/m minimum. In this case, there is a risk that the product might not pass even 2 V/m once measurement uncertainty is factored in.

To ensure that a product passes the test levels, a simple—but not necessarily cost-effective—solution is to start by overtesting a product by a factor of at least 4 dB. This means that a product required to pass 3 V/m should be tested to about 5 V/m, and a product required to meet the 10-V/m requirement should be tested at about 16 V/m. If a product fails to pass the higher levels, then it might be appropriate to examine the laboratory's uncertainty calculations and test-level calibration to ensure that the products are meeting the NIST-traceable levels without having to overtest them.

Compensation for Amplitude Modulation

The standard specifies that a continuous wave (CW) field must be used for calibration of the uniform field. Later, during testing of an actual product, 80% 1 kHz sine-wave amplitude modulation is typically used to further simulate the disturbance presented by a VHF or UHF communications device. The same forward power must be used for calibration and during the actual test.

Although analog communications theory is rapidly becoming outdated, many engineers might recall from their core studies that the introduction of a modulating signal introduces additional power into the system. For the case of amplitude modulation at 80% depth, the additional power introduced in the sidebands is 1.206 dB. Therefore, during the performance of the IEC 61000-4-3 test, the forward power level should be reduced by this amount to maintain the forward power used for calibration.

Monitoring During Testing

Since the IEC 61000-4-3 test uses a precalibration to measure the performance of subsequent testing, how a laboratory monitors for the appropriate test level is critical to establishing the correct calibrated field level. The best approach is to use a directional coupler and power meter to monitor forward power at the output of the power amplifier. This ensures that all necessary system components (including the amplifier, cabling, and antenna) are working properly to establish the proper forward power to the test system.

In some cases, a laboratory may use the voltage output of the signal generator as the reference. Although this may function properly as a reference point under ideal circumstances, this method does not monitor for the proper operation of the rest of the system, such as the power amplifier gain or antenna and cable connections.

Conclusion

Because of the number and complexity of tests that must now be performed, many manufacturers rely on the services of outside laboratories for expensive and complicated tests such as those described in IEC 61000-4-3 for immunity to radiated electromagnetic fields. Laboratories must often interpret the intent or requirements of the standards. Familiarity with the test standard and its requirements helps to ensure that laboratories provide results that are in line with a company's own goals and objectives.

Clark Vitek is an EMC staff engineer for CKC Laboratories Inc. at its Hillsboro, OR, location.

Back to the Fall 99 Table of Contents