Free-Space Chambers for Radiated Emissions Measurements

Free-space chambers have some very real possibilities for revolutionizing compliance testing. Along with supplying the theoretical equivalence to measurements made on an open-field site, they could be an economical, efficient and accurate answer that benefits both manufacturers and regulatory agencies.

The free-space chamber is a completely RF absorber-lined shielded room designed to optimize field uniformity during the performance of radiated RF immunity testing. This type of chamber also generates interest for performing FCC/ANSI C63.4 and CISPR 22 (1993) type radiated-emissions evaluations.

The free-space chamber allows a single compact, relatively low-cost facility to test for both radiated immunity and radiated emissions. These chambers simulate free space by placing RF anechoic material on all surfaces of the test enclosure, including the floor.

For this reason, the chambers are sometimes referred to as completely absorber-lined chambers or fully anechoic chambers. These chambers differ from other types of radiated-emissions test facilities, such as an open-area test site (OATS) or a semi-anechoic chamber. In free-space chambers, radiated emissions can be accurately measured without a ground reference plane as part of the measurement system.

Comparison to Other Facilities

By eliminating the ground reference plane from the measurement system, the maximum emissions from the equipment under test (EUT) can be obtained at a fixed antenna height in a free-space chamber. This obviates the need to vary the scan antenna height required to achieve similar accuracy at an OATS or in a semi-anechoic chamber.

Typically, a 1.5-meter fixed antenna height in the free-space chamber can yield equivalent results to a semi-anechoic chamber where the receive antenna height must be varied from 1 to 4 meters. To accommodate this scan height, a semi-anechoic chamber must have an inside height of approximately 5 to 6 meters, compared to the fixed 3-meter inside height typical of a free-space chamber design.

Eliminating the need to vary the scan antenna height speeds up emissions testing. It also improves the accuracy and repeatability of measurements by removing scan height as a test variable.

Because the free-space chamber is a shielded indoor facility, it also shares a major advantage with semi-anechoic chambers over open-field testing: the capability to obtain emissions measurements in an ambient-free environment. This eliminates problems often encountered at OATSs with local transmitters such as television and radio stations interfering with the test. It also reduces the possibility that emissions from the EUT will be overlooked.

Because free-space chambers can be constructed with approximately one-half of the typical inside height of a semi-anechoic chamber, they are substantially less expensive to build. They also can fit into a standard commercial building with about 12 ft of clear height.

Chamber Construction

A typical free-space chamber measures 16’ × 30’ × 10’ and is lined with ferrite tile, a pyramidal absorber or a combination of these different types of RF anechoic materials. The absorber materials should provide approximately -20 dB of return loss over the desired test-frequency range, usually 30 MHz to 1 GHz. Higher frequency materials can extend the usable range to 40 GHz.

Because the floor is lined with RF absorber, a raised wooden floor usually is installed to protect the absorber. The raised floor accommodates an EUT turntable. The raised floor and turntable must be constructed entirely of nonconducting materials to preserve the free-space condition.

Figure 1 shows a typical ferrite tile-lined free-space chamber. The overall length and width depend primarily on the size of the EUT, the type of antennas used and the desired measurement distance. In general, the room should maintain approximately 1 meter to the absorber materials from the largest anticipated EUT and from the edges of the measurement antennas (in both polarities).

A typical room sizing diagram is shown in Figure 2. The cost of constructing a free-space chamber in a standard suburban commercial office building ranges from approximately $150,000 for a small chamber (5’ dia turntable, 3-meter test distance) to approximately $300,000 for a larger chamber (larger turntable or a 10-meter test distance). This compares to an equivalently capable semi-anechoic chamber costing approximately $200,000 to $400,000.

The actual cost of installing a semi-anechoic chamber is significantly greater because a building to accommodate a semi-anechoic chamber also must have a large inside clear height, usually more than 6 meters. This usually requires a special building to house the chamber, sometimes adding several hundred thousand dollars to the typical project cost (Table 1).

Radiated-Emissions Measurements

Methods of correlating free-space measurements to their equivalent measurements on OATSs generally show excellent agreement when the models are used to equate free-space measurements to open-field measurements.^1,2,3,4 A method of correlating free-space measurements to their open-field equivalents is necessary because all international specifications for limiting radiated emissions are based on measurements made over a ground plane.

Before measuring emissions, the free-space chamber must truly represent a free-space condition within reasonable tolerances (±4 dB of ideal free space). The equations cited in this article provide a theoretical model to validate a newly constructed free-space chamber and the basis for obtaining correlation factors to correct free-space measurements to their equivalents on an open-field site.

Site-Attenuation Model

A method for validating the construction of free-space chambers through normalized site attenuation (NSA) measurements proceeds similarly to the validations required for OATSs.¹ ANSI C63.4 and CISPR 22 contain tables of theoretical NSA for validation of an OATS or semi-anechoic chamber, but do not include any standard models for free-space facilities.^5,6 A general expression for NSA in free space, presented by Roger McConnell, can validate a free-space chamber:¹

where: V_direct is the measurement cable reference.

V_site is the antenna to antenna measured voltage on a receiver or spectrum analyzer.

AF₁ and AF₂ are the antenna factors of the transmit and receive antennas.

The left-hand side of this expression is the equation for NSA as it appears in ANSI C63.4 for normalized site-attenuation measurements using broadband antennas. The terms V_direct, V_site, AF₁ and AF₂ have the same definitions as in ANSI C63.4.

The right-hand side of the equation expresses free-space NSA in terms of only the source and load resistances, R_s and R_L of the signal generation and measurement equipment (usually 50 W ), wavelength l in meters, measurement distance D in meters, and free-space impedance , Z_o, which is 120 p W .

Measurement of NSA in accordance with the procedure described in ANSI C63.4, paragraph 5.4.2, for an alternate test site, may then proceed as usual for the free-space chamber—with one exception. The receive-antenna height should remain fixed at the center of the free-space chamber (1.5 meters for a normal 3-meter interior chamber height). The expression for free-space NSA should replace theoretical NSA values provided in ANSI C63.4.

Equation 1 for NSA is valid for a semi-anechoic chamber or an OATS through the use of a second set of equations to obtain the effective free-space distance for facilities with a ground reference plane. These equations substitute an effective free- space distance, D_eff, for the free-space measurement distance, D, contained in Equation 1.

The introduction of a ground plane into the NSA model requires that the NSA be treated separately for vertical or horizontal antenna polarity. For vertical polarity, the effective free-space measurement distance is:

For horizontal polarity:

where: D₁ and D₂ are the direct and indirect path lengths.

b is 2p /l .

By setting D_eff = D and substituting into Equation 1, a uniform model for NSA for a free-space chamber or a facility with a ground plane (such as an OATS or a semi-anechoic chamber) is obtained. This effective distance model appears in slightly different form and produces nearly identical theoretical NSA values as they appear in ANSI C63.4 (1992) Tables 1 and 2 for OATS and semi-anechoic chambers. ⁷

The three equations for NSA provide a single set of equations to obtain theoretical NSA values for any test distance, frequency, transmit and receive-antenna height. It includes free-space chambers as well as facilities with ground reference planes in the model.

A sample NSA calibration of a free-space chamber (Figure 1) is shown in Figure 3. The facility is within ±4 dB of the theoretical model of Equation 1. Similar NSA results have been obtained for chambers varying in size from 10’ × 24’ × 10’ to 20’ × 30’ × 10’.

Correction Factors

From the NSA equivalence between free-space chambers and OATSs, the concept of effective distance also can be used to obtain the necessary correction factors to correlate measurements made in a free-space facility to their equivalents on an open-field site. The correction factors are obtained by:

D NSA = 20 log (D_eff /D) (4)

where: D_eff is the minimum effective distance found when varying the receive antenna on the OATS (usually 1 to 4 meters).

D is the measurement distance in the free-space chamber.

By computing D_eff for any open-field test distance (such as 10 meters for the CISPR 22 Class A or B limits) and using the free-space chamber measurement distance (for example, 3 meters for the chamber shown in Figure 1), it is possible to correlate free- space measurements made at any test distance to their equivalents on the OATS. The expression D NSA must be computed separately for vertical and horizontal antenna polarity and at each frequency of interest. The value of D NSA (in dB) is then added to the measurement (in dBµV/m) obtained in the free-space chamber for comparison to the appropriate specification limit on the OATS (also in dBµV/m). Table 2 shows data obtained for a sample ANSI C63.4 paragraph 11-type tabletop setup (including CPU, monitor, mouse, keyboard and printer) measured in a 3-meter free-space chamber and then on a 10-meter OATS.

Limitations

Correlating free-space measurements to OATS equivalents has a primary limitation: At lower frequencies (30 MHz to 80 MHz), measurements of an actual EUT may be affected differently by the presence of a ground plane. This is of particular concern for floor-standing equipment normally located close to the ground plane during testing in OATSs. As a result, for floor-standing equipment and large equipment in general, more research into the limitations of these techniques is needed before the methods are applied to all equipment.

By using a fixed antenna height, some EUTs that exhibit a nonisotropic field pattern (directional gain) will measure differently when the scan antenna height varies. This creates some uncertainty in the free-space-to-open-field correction based on an isotropic model.

This limitation can be somewhat mitigated by testing at a closer distance in free space than on an open-field site; for example, a 3-meter free space compared to a 10-meter open field site. A comparison of the angle formed by an antenna 0.7 meter above the source at a 3-meter test distance in free space to a 4-meter-high receive antenna on an open-field site at a 10-meter test distance is within 0.08º.

Conclusions

Free-space chambers offer a low-cost alternative to testing in an OATS or in a semi-anechoic chamber. Free-space chambers have a theoretical equivalence to measurements made in an open-field site. But most important, they have the capabilities to test more quickly at a fixed antenna height and in an environment completely protected from interference by local ambients.

Because the free-space chamber is protected from local ambients, these chambers can be constructed in locations close to metro areas. This provides convenience and a lower overall cost for manufacturers who need to perform emissions testing. The acceptance of free-space measurement facilities also will aid regulatory agencies in their efforts to encourage manufacturers to meet the international standards for preventing electromagnetic interference.

Although semi-anechoic chambers share these same conveniences and test environment benefits over OATSs, the cost of constructing full scan height semi-anechoic chambers has limited their popularity. The relatively low-cost alternative of free-space chambers will encourage most manufacturers of medium to high volume products to construct their own facilities. Independent test laboratories can better serve their markets with these more efficient and cost-effective test facilities.

References

1. McConnell, R.A. and Vitek, C., “Calibration of Fully Anechoic Rooms and Correlation with OATS Measurements,” IEEE 1996 International Symposium on Electromagnetic Compatibility, August 1996, pp. 134-139, Santa Clara, CA.

2. Ristau, D. and Hansen, D., “Correlation of Fully Anechoic to OATS Measurements,” ITEM Update 1996, R&B Enterprises, pp. 78-83.

3. Leferink, F.B.J., Groot-Boerle, D.J. and Puylaert, B.R.M., “OATS Emission Data Compared with Free Space Emission Data,” IEEE 1995 International Symposium on Electromagnetic Compatibility, August 1995, pp. 333-337, Atlanta, GA.

4. Garn, H., “Radiated Emission Measurements in Completely Absorber Lined Anechoic Chambers without Groundplanes,” IEEE 1989 International Symposium on Electromagnetic Compatibility, pp. 390-393, Denver, CO.

5. ANSI C63.4, “American National Standard for Methods of Measurement of Radio-Noise Emissions from Low-Voltage Electrical and Electronic Equipment in the Range of 9 kHz to 40 GHz,” 1992, Institute of Electrical and Electronics Engineers, ISBN 1-55937-212-5.

6. CISPR 22, “Limits and Methods of Measurement of Radio Disturbance Characteristics of Information Technology Equipment, 2nd Edition,” 1993, International Electrotechnical Commission, Geneva, Switzerland.

7. Smith, A.A., German, R.F. and Pate, J.B., “Calculation of Site Attenuation from Antenna Factors,” IEEE Transactions on Electromagnetic Compatibility, Volume EMC-24, No. 3, Aug. 1982.

About the Author

Clark Vitek joined CKC Laboratories in 1992 and today is an EMC engineer at the company. He has presented papers at the IEEE International Symposium on EMC in 1987 and 1996. Mr. Vitek earned a B.S.E.E. degree from the University of California. CKC Laboratories, 5289 N.E. Elam Young Parkway, Suite G900, Hillsboro, OR 97124, (503) 693-3543.

Table 1.

Table 2.

Copyright 1997 Nelson Publishing Inc.

May 1997