Finding a day of test time at a commercial test laboratory has never been easy. Now, it can be nearly impossible since EMC test labs are filled to capacity and beyond.
Money also is a factor. The actual cost per shift is increasing along with the wait. Recently, the lead times for testing have reached as long as five months at commercial labs. As a result, two major trends have developed:
Competent test laboratories that meet all necessary technical requirements are being used only for compliance demonstration testing (CDT). This minimizes the test time required for CDT.
Many manufacturers are performing precertification testing on their products to ensure successful CDT.
These trends are currently defining the typical practice for developing and demonstrating compliance with all appropriate EMC requirements.
Best EMC Test Approach
The trend toward separate precertification testing as preparation for CDT is viewed by many as one of the best approaches to EMC compliance. However, many companies do not want the cost or the responsibility of operating a complex EMC compliance demonstration laboratory.
Instead, they buy a minimal set of equipment for radiated emissions testing and invest in the time to perform measurements that show compliance in a less structured environment than that of a compliance demonstration laboratory. They also incorrectly assume the equipment under test (EUT) that meets radiated emissions requirements is likely to meet radiated immunity requirements.
Often, precertification testing is performed in a makeshift radiated emissions test facility, under sometimes unusual conditions. They hope to reduce the six (or maybe 10) highest emissions to values nominally below the desired specification limit to ready the EUT to meet all emissions and immunity EMC requirements.
Unfortunately, this may not always be the case. There are two technical issues to address: radiated testing and conducted testing. For radiated testing, there is no guarantee that if the EUT meets all emissions requirements it will meet the immunity requirements.
Consider that the electromagnetic shielding values derived from an equipment chassis are a function of source impedance and separation distance from the source to the shielding barrier. Generally, meeting radiated emission specifications requires higher levels of shielding in decibels than meeting radiated immunity requirements. Since internal sources are much closer to the shielding barrier than external sources (at 3 m), shielding from inside sources is more difficult to achieve.
The conditions for internal vs external source shielding are summarized in Table 1. It indicates that higher shielding values are required to comply with radiated emission limits than with immunity limits. For radiated emissions, you must reduce the level from volts in the circuits (millivolts/m) in the chassis to mV/m at 3-m or 10-m distance, requiring about 60 dB of isolation. For radiated immunity, you need V/m at the surface of the EUT reduced to tens or hundreds of mV/m or about 40 dB.
For conducted emissions, the mismatch in requirements is even larger. Conducted emissions limits are low, even very low, when compared to the level of immunity test signals that will be applied. This suggests that the proper input power filter will not necessarily present symmetry from input to output as from output to input.
The analysis does suggest that if the proper design approach is taken for passing conducted and radiated emissions requirements, then the stage is set for meeting conducted and radiated immunity requirements.
EMC Precertification Testing
By running precertification radiated emissions tests, you can usually improve EUT performance, at least when a set of preliminary measurements is available from a facility approaching compliance demonstration quality. In this case, you can reduce the measured values of the emissions that exceed or approach the specification limits.
When the precertification measurements are not carried out in measurement conditions very similar to compliance demonstration facilities, then the visit to the compliance demonstration laboratory can be devastating. Even with the best intent and careful adherence to the measurement practice, conditions for the precertification measurements could vary significantly from the formal test, providing very different data than expected.
Measurements performed, either with or without a ground plane with EMC measurement instruments of marginal quality, will not produce the same results as obtained at a formal test facility. So unless careful attention is paid to the measurement equipment, the test facility and the measurement instrumentation, precertification testing can be an exercise in futility.
Often, inexpensive measurement instrumentation with excessive measurement errors is used for the precertification emissions measurement. In addition, uncalibrated antennas or antennas never intended for measurement use are utilized for precertification measurements at convenient distances, often less than the required measurement distances.
The measurement environment is often without a ground plane to stabilize reflections. Under these circumstances, data produced can be both highly variable and substantially lower than would be measured at a formal test site. To minimize these problems, use quality test equipment and base the precertification measurements on a set of data from a controlled environment.
Measurement Uncertainty Requirements
The newest issue at EMC test laboratories is a requirement to include uncertainty values for each measurement. It is likely that European emission standards will include an allowable uncertainty value, UL, for each emission specification limit. UL will be added to the specification limit.
To show compliance, the measurement uncertainty of the test setup, U, must be compared to the published uncertainty value UL. U can be smaller than the published value, U< UL, or larger than it, U> UL. While the uncertainty values are generally both positive and negative, only positive values increase the measurement value and they are the values to be considered.
Several conditions of measurement and uncertainty are shown in Figure 1. As shown in the figure, several interpretations are possible. The specification limit and a specification limit plus the allowed uncertainty of measurement also are shown.
Unequivocal compliance occurs when the measured value is less than the specification limit and the measurement uncertainty is below the specification limit. Equivocal compliance occurs when the measurement is below the specification limit but the value associated with the measurement plus the uncertainty is above the specification limit. Both will be regarded as an “accept” for the individual measurements.
One of the largest contributors to overall measurement uncertainty, U, is the individual uncertainty values associated with the antenna, items usually beyond the management of precertification setups. These items include, but are not limited to, antenna calibration, height of measurement uncertainty, VSWR, phase center variation and interpolation for values between measured points.
For a specific antenna, the individual uncertainty value, ui, is approximately ±2.55 dB. Overall measurement uncertainty for the measurement, U, is given by U=2uc, where uc =Precertification Testing Before CDT Is Sound
The growing strategy of using precertification tests to prepare for compliance demonstration testing is basically sound. The measurements should be accomplished on a relative basis using a reference set of data obtained under the best possible conditions, usually a compliance demonstration laboratory.
The precertification testing must be performed as closely to standards requirements and must be almost as strict as those for CDT. Failure to do so leaves the practitioner all but stranded in the sea of compliance without a life preserver.
The addition of measurement uncertainty to precertification testing means measured signals must be suppressed by an additional amount of the uncertainty value to ensure compliance with specification limits during CDT. This, in turn, implies additional cost for suppression which must be applied to each unit to ensure uniform performance.
The additional cost of suppression components may only be $1 to $2. If large production runs of the product are anticipated, the additional suppression costs can easily be more than the cost of high-quality test equipment.
Conclusions
The admittedly expensive acquisition of quality EMC test equipment can be a wise investment. By performing accurate precertification tests, you can reduce potentially unnecessary EMC suppression costs, which can save a significant amount of money throughout the product’s life cycle.
More importantly, the increased precertification test accuracy will help to ensure that the product meets all emissions requirements under all formal conditions of testing. It is a false economy to purchase inexpensive test equipment, particularly antennas, when the operating parameters and associated uncertainty values are unknown and difficult to estimate. This is particularly true when the penalty for an inaccurate estimate is so costly.
About the Author
John D. M. Osburn, the principal EMC scientist at EMC Test Systems, has more than 30 years experience as an EMC engineer. He has a B.S.E.E. degree from the University of Texas and a master’s degree in systems management from the University of Southern California. Osburn is a National Association Radio and Telecommunications Engineers Certified EMC Engineer and a member of the IEEE EMC Society Standards Committee. EMC Test Systems, 2205 Kramer Lane, Austin, TX 78758-4047, (512) 835-4684.
Table 1.
Source to Barrier Distance
Source Impedance
Radiated Emissions
Short, from the inside of the chassis to the chassis is only fractions of a meter
Varies from power distribution circuits (very low) to signal circuits (low to high), usually near- field conditions
Radiated Immunity
Typically three meters, sometimes closer
377 W , the impedance of free space; typical for far- field conditions
Figure 1.
Copyright 1997 Nelson Publishing Inc.
February 1997