What’s Behind EMC Test Software?

Explore the behind-the-scenes information that software must have to ensure that testing is done properly.

Electronic and electrical equipment manufac-turers are being placed under more and more scrutiny to ensure their products meet national and international electromagnetic requirements from emissions to immunity. As a result of this scrutiny, it is increasingly important that laboratories evaluating products demonstrate that their test software provides accurate results to help ensure compliance with standards requirements.

Immunity software for radiated and conducted susceptibility also plays an important role in ensuring the performance of a product in the presence of RF interference. If the software does not properly record calibration numbers or repeat the necessary power levels to generate the required test conditions, products may be under- or over-tested, providing false results costing manufacturers money and time to market.

Software is both a blessing and a curse. It is a blessing for the speed and consistency it can bring to bear on complex measurements. However, it can be a curse since, by its very nature, it masks those same complexities, creating opportunities for errors and omissions that can compromise measurement accuracy.

Emissions Software
The most common software packages used in interference testing address precompliance or full compliance testing. In either package, the basics of emissions testing not only must be understood by the writer, but also must be easily communicated and customized.

The typical information generated by emissions software can be seen in Figure 1. It shows graphical and tabular results of a product tested. But what is required in a software package to get to this final result?

Figure 1. Typical Graphical and Tabular Data Reporting

The capability to customize parameters, such as frequency ranges, transducer and gain/loss values, and resolution and video bandwidths, must be provided by measurement software. The software also should allow you to customize the way in which measurements are taken in terms of detectors, sweep rates, and input selections.

Typical detectors used in EMI measurements today include, peak, quasi-peak, average, and rms. Each detector is defined in the CISPR 16 series of standards and has minimum required bandwidths and sweep rates to ensure measurements are taken accurately.

Software should allow for selection of a variety of spectrum analyzers or EMI receivers, preselectors, detection adapters, turntables, switch drivers, and signal generators or tracking generators. Control of such devices demands that the software has all the necessary information about the equipment it controls to ensure that you can appropriately configure the instrument.

Take instrument bandwidths as an example. Certain measuring devices have the capability to measure in both 3-dB and 6-dB bandwidths. Without the capability to select the appropriate bandwidth for the standards used as a measurement basis, the maximum emission may not be measured. This is especially true if the standard requires a 6-dB bandwidth and the software only allows for a 3-dB bandwidth to be chosen.

Measurement Accuracy
When configuring EMI software, you should have the capability to improve the accuracy of the measurements that will be taken. For example, if a log periodic dipole array antenna is used to measure between 200 MHz and 1,000 MHz, depending on the measurement device, the frequency resolution could be between 500 kHz and 2 MHz. The number of points a measuring device can make while sweeping dictates the resolution.

Software should have the capability to subdivide the measurement range to provide better resolution. The same 200-MHz to 1,000-MHz frequency range broken into four subranges, each 200 MHz wide, now will provide a resolution of 500 kHz, which allows much more accurate final results. This technique also will reduce the possibility of not measuring incorrect frequencies or voltages at particular frequencies that may exceed the limit in the final detector.

It is important to recognize these two pitfalls. When the software reports the peak emissions, the values of the frequencies may have as much error as one-half the resolution, meaning that measurements in the final detector would not be the maximum voltage emitting from the product under test. For the single range example, this would mean there could be as much as a 1-MHz frequency error.

Depending on the measuring device, the more resolution points the analyzer provides, the less subranges would be required. Remember, too, that the more subranges you use, the slower the entire measurement process will be. So in cases where peak scans are used for preliminary results prior to obtaining the final quasipeak or average measurement, good engineering judgment must be used so accurate results are obtained in a reasonable amount of time.

The sweep time per range is another important parameter to consider when defining scanning ranges. To ensure the maximums of all signals, whether transient or continuous, are measured, the software must accommodate adjustable sweep times. If it does not, it is possible to sweep too fast and miss the maximum of any transient signal.

Error Checking
Emissions software must provide error-checking capability. It not only compares the selected parameters to the common standards requirements, but also ensures that, based on the specific measuring device, the selected parameter is allowed.

Figure 2 shows an example where the programmed resolution bandwidth was defined as 100 kHz for the 30-MHz to 1,000-MHz range. For quasipeak measurements, the normal bandwidth in this range is 120 kHz per CISPR 16-1-1. If the software did not detect this, then you would be making measurements using a final detector that would not provide accurate or correct results.

Figure 2. Example of a Setup Error Detection

Measurement Control
After the basics are programmed into an execution routine, the software also should accommodate many different measurement options. Manual as well as automated measurements should be available.

Conducted Emissions
For conducted emissions on power lines or I/O lines, the software must perform standard quasipeak measurements over a defined frequency range or at a discrete frequency. Average measurements for conducted emissions sometimes are required, and the software should perform the necessary measurement including switching between log and linear magnitude references on the measuring instrumentation.

Radiated Emissions
For radiated emissions, the software must control the antenna mast tower and turntable. For prescanning or preliminary investigations, the software should be configurable to accommodate large steps between antenna heights and continuous rotation of the turntable while holding the measuring device�s trace in Max. Hold.

Depending on the source of the emissions, it may be necessary to use a fixed increment to rotate the turntable based on a duty cycle or repetition rate to ensure measurement of the maximum amplitude. The software should have provisions to select and define the azimuth control.

Final Radiated Emissions
Final radiated emissions measurements require the software to control the mast and turntable so the maximum signal received is recorded. If final measurements are made automatically, the software should prompt you to set up specific parameters to ensure that the frequency and amplitude are measured accurately during the process.

The software requires you to define the initial search span, a correlation check between the original peak scan, the measurement, and if necessary, the capability to remeasure the signal if it is within a specified limit margin. All these parameters must be controlled during automated measurements, guaranteeing that the proper signals are accurately tested. This is particularly useful when dealing with ambient signals at an open area test site.

While automated measurements are one method of getting the final results, you may prefer to control the equipment manually and have the software acquire the data from the test equipment and process the results by adding any correction factors Figure 3.

Figure 2. Example of a Manual Control Options

Immunity Software
Immunity software can control equipment for both radiated and conducted susceptibility testing. The primary function of this software is to drive a transducer to a specified level and control the level for repeatability. As a result, observations may be made relating to the performance of the product in the presence of RF signals rather than collecting data to determine product compliance as in the emissions software. It is a useful feature if the software can monitor a test point on the product and determine if its value is within a specified tolerance.

Behind the scenes, this software also is responsible for test-equipment protection and data manipulation. Immunity software should address test methods, frequency ranges, and equipment control and protection. The protection feature is important since amplifiers provide high levels of RF power that, if not controlled, can damage antennas, attenuators, and coupling devices.

Basic parameters should be configurable prior to calibration of field uniformity for radiated immunity testing or calibration of coupling/decoupling networks (CDN) or injection clamps for conducted immunity. Frequency range, coupling factors for direction/bidirectional couplers, and field probe control are just a few of the critical parameters that must be controlled.

Test Configuration
Radiated immunity in accordance with IEC 61000-4-3:2002 requires a uniform field to be measured. As part of the evaluation, the software must compare the data at each location in a standard calibration grid and determine its acceptability.

The standard requires that 75% of all points are within 0 to +6 dB of each other. An additional 3% of the frequencies can be as high as +10 dB. If the software cannot properly evaluate this information, the test levels applied by the lab may not comply, and the product under test may not be properly evaluated. Remember that the more uniform the field, the less over-testing and subsequent design costs are incurred.

For conducted immunity, the calibration is not as complicated, and the software is not as labor-intensive since only one test level is calibrated into either a CDN or into an injection clamp (EM-clamp or bulk current injection probe). However, it is important that the software be able to correct for impedance changes.

The CDN has an input impedance of 150 ?, and typical measuring instruments have 50-? inputs. Due to this impedance mismatch, the software should be able to account for the voltage division that occurs and correct the drive level to ensure the proper calibration level is achieved.

As part of the overall test configuration, the software must control the amplifier by switching from standby to operate and back when needed. It also should record the forward and reverse power levels delivered to the appropriate transducer and control the sweep times, modulation levels and types, and the desired test level.

For both radiated and conducted immunity, the software must properly address the handling of signal input power to the amplifier. Many signal generators create switching transients from internal attenuators and switches.

Figure 4 illustrates a transient occurring between -14 dBm and -15 dBm. This example is a 4.7-dB transient. Transients can adversely affect products under test, especially at high field strengths, giving a false indication of noncompliance.

Figure 4. Example of a Signal Generator Transient

Transients of this nature also can damage or degrade amplifiers. Programmers can avoid this problem through detailed knowledge of the instruments and adjust the algorithm accordingly.

Safe Drive Levels and Available System Power
Since amplifiers are a source of high-level RF signals, it is important for the software to include constantly monitored information to ensure the entire immunity system is protected against damage.

Amplifier-safe drive levels are an important parameter that RF immunity software must accommodate as part of the system configuration. Specified by the manufacturer, this is the highest applied level that may be applied from the signal source before the amplifier goes into saturation and creates harmonics that may influence the test. Safe drive levels ensure that the amplifier will not be damaged during operation.

Available system power is another important parameter that immunity software must address. This is the amount of power the system needs to produce the desired test level.

Two power levels must be considered:

� Unmodulated Power

If you select a level during calibration that exceeds the available power that can be delivered by the system, the software should warn you of the possible dangers of continuing or abort the test and provide a message explaining why the test was aborted.

� Modulated Signal Level

Calibrations typically are performed prior to applying modulation. After calibrations are performed and a test is run, the modulation is applied.

Depending on the modulation type, the system may need to apply as much as 1.8 times the initial unmodulated power or 5.1 dB more power than required for a continuous wave signal. The software should be able to quickly determine whether or not the system can deliver the required field strength if modulation is added to the calibration level. This, of course, is based on the net power delivered during the calibration.

Test Flexibility
Once a calibration is performed and testing is ready to begin, the software should have built-in flexibility to provide test-parameter inputs. These inputs include the test frequency range, dwell times, modulation characteristics, and EUT-specific information such as internal clock frequencies that must be investigated as part of most immunity standards.

The software also should allow for the input of specific critical frequencies required by the standards. For instance, Telcordia�s GR-1089-CORE, Electromagnetic Compatibility and Electrical Safety�Generic Criteria for Network Telecommunications Equipment specifies that a list of key frequencies be investigated to ensure that equipment on the telecommunications network is not affected by services such as cellular telephones, personal communications systems, and certain broadcast television stations, to name a few.

Immunity software must monitor the system power during test. To account for the effects of impedance mismatches in the case of conducted immunity or signal reflection in the case of radiated immunity, the software should be able to ensure the test signal is not affected to the point where it is too low or too high during the test. The capability to monitor and control the power delivered to the appropriate transducer during test will ensure the product is subjected to the required levels.

For instance, when testing signal lines for conducted immunity in accordance with IEC 61000-4-6: 2001, a feedback system is required to monitor the induced currents on the lead being tested. This will ensure the induced currents are not significantly high if the impedance of the line is much lower than 50 ?. If the level is greater than 6 dB above the calibration level, the software should lower the signal generator drive level until the induced current is no more than 6 dB above the calibration level.

Immunity software also should monitor test voltages from the EUT. This will allow you to define operational limits that can be used as pass/fail criteria during testing. If the monitored voltage is outside the preset ranges, the software indicates that the product has deviated from the expected values.

A plus in radiated immunity software is the capability to optimize the field uniformity. This is achieved by varying the height of the transmit antenna. By achieving a more uniform field, there is less chance of subjecting a product to higher field strengths than required.

Allowing a 0 to +6-dB variation could double the voltage seen by the product at various points in the calibration grid. By varying the antenna height, the variations seen across the grid can be reduced significantly and ensure the product is tested to the levels required.

Conclusion
The software you choose should provide the capability to increase test accuracy and productivity and ensure testing is conducted in accordance with standards requirements by checking test configurations and equipment settings.

Some software packages will support equipment from multiple vendors while others will only work with one or two. When approaching software suppliers, let them know if the software will need to support currently owned assets to ensure compatibility. Ask for demonstrations, and take advantage of available training. This will help you make the right selection. Do not let a friendly user interface mask potential problems or poor algorithms.

References
Telcordia Technologies Generic Requirements, GR�1089�CORE, Electromagnetic Compatibility and Electrical Safety�Generic Criteria for Network Telecommunications Equipment, Issue 3, October 2002.

ANSI C3.4, American National Stand-ard for Methods of Measurement of Radio-Noise Emissions from Low-Voltage Electrical and Electronic Equipment in the Range of 9 kHz to 40 GHz�Revision of ANSI C63.4-2001, Jan. 1, 2003.

IEC 61000-4-3, Electromagnetic Compatibility (EMC) Part 4-3: Testing and Measurement Techniques�Radiated, Radio-Frequency, Electromagnetic Field Immunity Test-Edition 2.1; Edition 2:2002 Consolidated with Amendment 1:2002, Sept. 1, 2002.

IEC 61000-4-6, Electromagnetic Compatibility (EMC) Part 4-6: Testing and Measurement Techniques Immunity to Conducted Disturbances, Induced by Radio-Frequency Fields-Edition 2.1; Edition 2:2003 consolidated with Amendment 1:2004, Nov. 1, 2004.

CISPR 16-1-1, Specification for Radio Disturbance and Immunity Measuring Apparatus and Methods Part 1-1: Radio Disturbance and Immunity Measuring Apparatus Measuring Apparatus-First Edition; Amendment 1: 06/2005; together with CISPR 16-1-2 thru CISPR 16-1-5:2003, replaces CISPR 16-1:2002, Nov. 1, 2003.

About the Author
Robert DeLisi is the principal engineer for EMC at Underwriters Laboratories. In 16 years with UL, he has managed and helped design some of UL�s EMC facilities. Mr. DeLisi participates on ANSI Standards Committee C63 and is the chair of Subcommittee 8, Working Group 2 and a member of CTL ETF 10 for EMC. Underwriters Laboratories, PDE Division�3615ESNK, 1285 Walt Whitman Rd., Melville, NY 11747, 631-546-2452, e-mail: [email protected]

October 2005

Sponsored Recommendations

Comments

To join the conversation, and become an exclusive member of Electronic Design, create an account today!