The Case for Early Testing

Before an electronic product can be sold, it must comply with EMC standards. In general, EMC specifications attempt to ensure that a product’s performance is compatible with its environment. Emitting signals that could disrupt operation of nearby equipment is not allowed. It’s also important that the product continues to operate correctly when subjected to emissions from other equipment.

The various power levels and frequency bands used to describe allowable performance are established in an effort to achieve compatibility. It is not the intention that there should be no EMI generated, but rather that it should be minimized. Obviously, emissions that could disrupt communications channels must be controlled, but there are many other areas such as industrial control circuits where excessive electrical noise also can cause problems.

Because of the well-documented large cost increase accompanying fault correction late in the design-manufacturing-distribution cycle, EMC problems must be identified as soon as possible. Precompliance testing is a means of finding EMC-related issues at the design prototype stage.

Although your design eventually will have to pass compliance testing, that procedure is different from precompliance testing. Both are based on the same national and international standards, but compliance testing must be done very accurately, on a properly prepared test site with equipment that complies with CISPR specifications. In contrast, precompliance testing is an investigatory process.

You don’t know if the product will meet the emissions or susceptibility limits so you need to find out. Testing approximates compliance test conditions, but rather than being accurate to 0.5 dB, it is intended to help you understand why some performance areas are marginal. At which frequencies and orientations is the product failing or close to failing?

Precompliance testing can be very interactive and iterative given the test- fail-troubleshoot-redesign-retest cycle inherent in prototype performance verification. Accordingly, ease of use has a different emphasis for precompliance test software than for compliance test applications. Total automation is much less important for precompliance work than having straightforward control of only the necessary functionality.

Precompliance Software Programs

Precompliance testing must be reasonably accurate but may be an infrequent activity in a small company. Test setups often will be ad hoc arrangements made without the benefit of a proper screened room, anechoic chamber, or open-air test site (OATS). And, engineers may lack detailed familiarity with test instruments as well as test specifications.

These are some of the factors Laplace Instruments’ products address. “Test-site calibration almost invariably is the major source of measurement error,” commented David Mawdsley, Laplace managing director. “Because they need a low-cost solution, engineers want a precompliance system that can be used in parking lots, loading bays, a spare conference room, or an odd corner of the lab. Of course, they have to deal with ambient signal interference in such an environment.”

The company’s precompliance test software is dedicated to the Laplace range of analyzers and, when used in conjunction with a calibrated reference source, will automatically measure the characteristics of any site. The software applies appropriate correction factors to make the site produce measurements approximating those that would be measured in a 3-m OATS. In addition, ambient cancellation effective under most conditions is provided.

The user selects the type of test being conducted, but the software does not allow other detailed inputs. Instead, it automatically sets variables such as resolution bandwidth, frequency step, and sweep rate to ensure that no measurement errors result from user inexperience or simple mistakes.

Further, a TestDirector button accesses an expert system to guide the user through the common CE directives leading to selection of the appropriate standard to suit the product and explanations of the test techniques and test parameters. A step-by-step demo is available from Laplace Instruments.¹

Rohde & Schwarz (R&S) markets ES-SCAN, a software tool suitable for all EMI tests prior to final certification. It is a 32-b Windows application that facilitates precompliance testing to commercial standards and works with the company’s ESPI3 and ESPI7 Test Receivers. Capabilities not required, such as remote control for turntables, antenna masts, or absorbing clamp slideways, are not included.

According to Karl-Heinz Weidner, product manager for EMI and signal analysis test and measurement products, “ES-SCAN provides quick and reliable recording, evaluation, and documentation of disturbance voltage, power, and radiation as well as storage of all test parameters and measurement data. Capabilities are tailored to the needs of development labs, and the user interface is clearly structured. The software only offers functions that are needed for precompliance or precertification testing.”

A predefined library of standard limit lines, transducer factors and settings, and scan tables simplifies test setup. ES-SCAN supports frequency scans with up to 1,000,000 measurement points for each of two traces. A large frequency range from 9 kHz to 7 GHz can be swept using two different detectors simultaneously. For example, you could use the new CISPR-Average (Amendment A1:2002) and RMS-Average (Amendment A2:2007) detectors available from R&S (Figure 1).

Click here to view Figure 1
Figure 1. Prescan Display Showing Peak and Average Measurements With Marked Local Maxima
Courtesy of Rohde & Schwarz

The benefits of ES-SCAN include the following:
• Ease of use and a short learning phase stemming from a clear structure and simple user interface.
• Reduced test time and increased measurement efficiency resulting from predefined standard measurement setups and quick programming.
• Reliable test reproducibility based on storage and management of all measured data, settings, and test parameters.
• Flexible and quick test-report generation.
• A built-in operation guide that prompts step-by-step procedures and avoids incorrect measurement settings.

As a result of a management buyout late in 2006, Schaffner’s Test Systems Division became Teseq. The new company specializes in EMC test solutions and has retained former Schaffner employees with extensive EMC testing experience.

Emipak 3 Precompliance RF Emission Software addresses both conducted and radiated forms. Based on the company’s well-established full-compliance software, Emipak 3 streamlines emissions testing. You can choose among a number of predefined tests that measure emissions relative to the limits you have selected.

Through a series of simple forms, you can customize test setups, input details of the EUT, select the frequency range, and make detailed changes to the type of detector, bandwidth, and dwell time. Failed points automatically are retested, and you also can add or delete points to be remeasured. The graphical and tabular data is formatted as a report, but both graphical and textual elements can be copied to a more detailed report written in Word.

Emipak 3 supports all Schaffner measuring receivers and a large number of Advantest spectrum analyzers. If you don’t have the relevant Advantest or Schaffner instruments, you can’t use Emipak 3. Similarly, unless you have an R&S receiver, ES-SCAN isn’t for you, and likewise, Laplace software requires Laplace instruments. Instead of these products, you may need a more generic precompliance software tool that works with a range of test equipment from different manufacturers.

The Model SW1006 Radiated Susceptibility, Conducted Immunity, and Emissions Test Software from AR Worldwide may be what you want. It was developed under the National Instruments LabVIEW environment that provides access to a library of more than 500 instrument drivers from 45 different manufacturers. Chances are that this program is compatible with your spectrum analyzer.

Capabilities include full compliance testing to several standards as well as the flexibility to select test parameters and change the test threshold for susceptibility testing. In addition, closed-loop leveling can be performed. The program is intended for precompliance emissions testing with a spectrum analyzer and either a preamp or LISN.

Similar to SW1006, EMITest also is a full compliance program that can be used for precompliance testing. EMITest was developed and is used by CKC Laboratories, an EMC test lab. According to the company website, because the software is a key part of the lab’s testing operations, revisions and improvements are constantly being introduced by CKC programmers. A long list of spectrum analyzers, signal generators, turntables, antenna masts, and GPIB interface cards is supported.

According to Randy Clark, an EMC engineer at the company, “EMITest can gather data from the spectrum analyzer in either tabular or graphical format. Using the built-in recall functions, it’s easy to configure a variety of common setups, allowing quick identification of known compliance problems. A list of frequencies can be added to the graphical data for easy comparison of clock frequencies or known trouble frequencies.

“The graphical data also can be compared with other data for A-B comparisons over a range of frequencies.” He explained, “This tool is especially helpful for seeing deviations from ambient conditions or picking out EUT vs. host emissions. Graphical data can be captured either automatically for simple configurations or semi-automatically, allowing signal manipulation while using the software to control the frequency ranges being investigated. When comparing data to various spec limits, the software can recalculate the data against a new spec limit with just a few clicks of the mouse.”

Although the recent change to a 6-GHz upper frequency obviously affected EMC test hardware, EMITest was not. Only a slight modification of the spec limit was needed. Using the EMITest spec limit editor, it is as simple as editing an Excel spreadsheet.

An extensive library of current test limits is essential, and EMITest has predefined limits for all the popular commercial emissions standards. A partial list includes EN 55011, EN 55022, FCC Part 15, EN 61000-4-3, and MIL-STD-461.

Preprototype Testing

If EMC testing at the engineering prototype stage is worthwhile, what about during design simulation? In addition to circuit models used to determine basic operation, the modeling process today includes thermal performance as well as EMC characteristics. The usefulness of different types of simulation depends on the level of detail you include in the various models.

At low frequencies, circuits comprising discrete components have almost ideal behavior. Raise the frequency, and you start to see deviations caused by stray capacitance and inductance not included in the original circuit simulation description.

Crosstalk and other layout-dependent effects also will become problems at higher frequencies in the actual circuitry. More complex component models and accurate layout information are necessary if software simulations are to exhibit the kinds of second-order effects seen in physical products.

Similar reasoning applies to EMC simulation software. EMI Analyst^™ is a precompliance design tool produced by EMI Software LLC. To determine a product’s EMI characteristics at the box level, calculations involve design parameters such as nonideal circuit components, cabling construction, wire routing and twisting, shield construction and termination, grounding, circuit impedance and balance, filtering, and field distributions and propagation.

EMI Analyst avoids repeating the detailed analyses associated with these many parameters by working at a much higher level. For example, engineers can use current and voltage time-domain waveforms to define the noise source.

The program’s designer and company founder Steve Newson explained: “It’s the voltage and current waveforms generated in a circuit that produce conducted and radiated emissions. We define the noise source by using these waveforms, which represent information readily available during the circuit design phase of the project.”

Four elements affect a circuit’s EMI performance: the source, any filtering that may be applied, the conductors connecting sections of circuitry, and the load at the end of the cable. The software’s built-in utilities help you define circuit parasitics, design suitable filters, describe wiring configurations, and calculate load impedances. An integral schematic capture capability accurately models the EMI effects caused by interaction of these factors.

The MicroStripes EMC analysis program from Flomerics accepts simulation data from several sources. For example, results can be based on a thermal model of the enclosure design. In this case, as explained by Paul Duxbury, a senior EM engineer at the company, “The simulation mimics a shielding effectiveness or radiated emissions test in a lab by calculating the electric field strength on a 3-m or 10-m radius cylinder around the model.

“Simulation not only determines the amount of radiation, but it also establishes the path by which it escapes from the enclosure. In a typical example, the simulation might show that most of the emissions escape through a particular seam. Placing gaskets along this one seam would increase the shielding effectiveness at a much lower cost than installing gaskets throughout the entire enclosure,” he concluded.

The behavior of complex sources is modeled by the compact source interface (CSI). MicroStripes works with the high-level data provided by CSI without the need to explicitly describe the details of a PCB, antenna, or noise source. Compact models also can be based on near-field data obtained from other PCB simulation and measurement tools. For example, Flomerics is collaborating with tool vendors such as SimLab and Sigrity to provide data-translation utilities to support the use of field data.

In contrast to other general-purpose EM simulation software, MicroStripes uses the transmission line matrix method of solving Maxwell’s equations. Mr. Duxbury said, “This method provides major advantages when performing EMC simulations because it solves for all frequencies of interest in a single calculation. As a result, it captures the full broadband response of the system in one simulation cycle. This characteristic is very useful in EMC analysis where potential resonances and emissions vary over a wide frequency spectrum” (Figure 2).

Summary

Precompliance EMC testing really does have objectives and constraints that are distinct from full compliance testing. Testing should be accomplished as economically as possible, but because precompliance tests may be repeated several times, cost can become a major factor. Also, test facilities typically will be very different for precompliance testing. Unless a company has a large number of products to test on a continuing basis, it can’t justify investing in a permanent EMC test facility and associated staff.

When deciding what software best suits your situation, it’s important to remember the reason that precompliance tests are run. Will the proposed software deliver the information you need to ensure that the product passes full compliance testing on the first try? That’s clearly the objective, whether you test conventionally with a physical prototype or via EMC simulation.

Reference

1. Laplace Instruments, http://www.laplace.co.uk/ERSCAL/sitecorr.htm

March 2008