If you want your products to meet radiated-emissions immunity requirements, they probably need some form of shielding. It may be localized shielding, such as a small shield around one or two components, or overall shielding of a system.
How Much Is Enough?
The level of shielding depends on the application. The higher the emission level, the higher the level of shielding required—and the more expensive the shielding. That’s why it’s important to determine the minimum shielding needed to isolate your product from the environment.
For example, the initial cost of an aluminum part is more than an impregnated or coated-plastic molded part and you may not need the amount of shielding supplied by the aluminum part. Experiment with several types and levels of shielding materials, considering both the level and the effective frequency range of each.
After choosing a shielding material, determine whether the shielding effectiveness is dependent on the production process. That is, can variations in the manufacturing process cause variations in the shielding effectiveness? If so, you may need to do sample or continuous testing.
Shields that are production process-dependent include spray-coated, impregnated and pressure/heat-activated materials. Spraying the interior of product covers with conductive (copper) paint is one way to contain EMI.
To ensure that the painting process is under control, measure each part or a sample of parts to determine the shielding effectiveness. One method measures the resistivity of the coating. The higher the resistance, the poorer the shielding. This method requires resistivity testing at several points on the sprayed surface.
A second method is easier and faster. A shielded cover is passed through a field and the reduction of the field is measured. This method directly indicates the shielding effectiveness of the part. The tester does not have to correlate resistivity to shielding effectiveness.
Establish a Reference Line
Before testing shielding effectiveness, establish a reference line. Place two probes facing each other at a distance far enough apart so that a shielding sample can pass between the probes without moving the probes (Figure 1). The reference line won’t be accurate if the probes are moved to accommodate the test sample.
Use a spectrum analyzer with a tracking generator for the reference measurement. The tracking generator supplies a signal source that is tuned to the same frequency as the spectrum analyzer.
Initial Equipment Setup
With the probes adjusted for the correct spacing, switch on the tracking generator and adjust the level to place the trace at the top of the screen. Figure 2 shows the top trace, which is without shielding between the probes.
To ensure there is enough dynamic range (the distance between the unshielded trace and the noise floor), switch off the tracking generator and view the noise floor. The minimum distance must be greater than the shielding required by the parts being measured (Figure 3). If the dynamic range is not sufficient:
Reduce the resolution bandwidth for better sensitivity (sweep speed slows).
Increase the power of the tracking generator (requires reference level changes).
Measurement Process
After the equipment is set up to meet the parameters of the measurement, perform a shielding evaluation. Most spectrum analyzers have more than one trace in which to write, store and view data. Up to three traces are available with the spectrum analyzer used in Figures 2 and 3.
Determine the relative shielding effectiveness using these steps:
Adjust the spectrum analyzer to the desired frequency range.
Without the shielding sample, take a sweep (tracking generator on) and write the results into the spectrum analyzer’s trace A.
Place the shielding sample between the probes. Take another
sweep and write the results into trace B.
Look at the difference between the two traces. If the vertical distance (in decibels) between the two traces is greater than the desired shielding, the part passes. If there is a point along the traces where the distance between the two traces is less than the desired shielding requirements, the part fails.
Automating Measurement
When the shielding is close to the specified limit, you may have a difficult time making the correct decision. If your spectrum analyzer has internal automation programming, removing operator subjectivity from the process is easy. A simple program can be developed which, when activated from the front panel, can take a sweep, compare the new trace to the reference trace, and flag the operator with a pass or fail displayed on the spectrum analyzer display (Figure 4).
More Automation
With the aid of parts-handling equipment, you can automate the process to a level where no operator intervention is required after initial setup. For this you need a spectrum analyzer that senses transistor-transistor logic levels on a line. The spectrum analyzer used here has a 9-pin bus connector with 4 output lines, an input line, 5 V and a ground. If this is not available, use a computer with an IEEE-488 bus.
Figure 5 illustrates this process:
The material handler positions the sample between the two probes and trips a photo cell.
The tripped photo cell is sensed by the input line of the
9-pin bus.
The internal program of the spectrum analyzer monitors the change on the input line and starts the measurement process as before in Figure 4.
After the test program determines whether the sample passes or fails, the levels on the output lines can be set to instruct the handler to accept or reject the sample and move another into place.
Conclusions
With the aid of a spectrum analyzer and a tracking generator, the relative shielding effectiveness can be measured easily. The level of automation depends on the number of units to be evaluated. In a laboratory setting, where the interpretation of the operator is required, a simple test may be satisfactory.
As the volume increases, it may be too costly to have a technically skilled operator perform a fairly routine task. Then, a program resident in the spectrum analyzer can help a less skilled operator perform at the same level as a technically skilled operator. For very high volumes, the fully automated approach may give you the best return on your investment.
About the Author
Dennis Handlon has developed EMC seminars, field sales training programs and products for the past five years at Hewlett-Packard. He joined the company in 1968 after graduating from San Jose State University with a B.S.E.E. degree. Hewlett-Packard Co., 1400 Fountain Grove Pkwy., Santa Rosa, CA 95401, (800) 452-4844.
Copyright 1995 Nelson Publishing Inc.
May 1995