A COMPOSITE NEAR-FIELD SCANNING ANTENNA RANGE FOR MILLIMETER-WAVE BANDS Doren W. Hess dhess@mi-technologies.com John McKenna jmckenna@mi-technologies.com MI-Technologies 1125 Satellite Boulevard Suite 1 Suwanee, Georgia, U.S.A. 324 Steve Nichols snichols@mi-technologies.com ABSTRACT This paper describes a Composite Near-Field Scanning Antenna Range for frequency bands that extend from X- Band in the microwave frequency regime through W- Band in the millimeter-wave regime i.e. 8.2 through 11 GHz. We show some of the initial checkout data using pyramidal standard gain horns and compare the patterns to theory. Keywords: Millimeter-Wave Measurements, Near-Field Scanning Range, Standard Gain Horn Measurements 1. Introduction A composite near-field scanning system is composed of two major positioning subsystems: an X-Y- planar scanner and a roll-over-azimuth spherical positioner. When integrated together, these form a near-field scanning system capable of three types of near-field measurement planar, cylindrical and spherical. A system of this type has been constructed and fielded for frequency bands that cover X-band through W-bands -- i.e. 8.2 GHz through 11 GHz. Here we discuss the operation at millimeter-wave frequencies. Even though the composite range approach is straightforward, the wide extent of the frequency-band incurs a need for special attention to numerous important factors, including: Positioning Accuracy Range Alignment Accuracy Truncation of the Measurement Surface Power Budget Dynamic Range and S/N Ratio In this paper we describe the approach taken to the construction of a composite range and show results based upon measurements of pyramidal horns that demonstrate the good performance achieved. 2. Descripton of the Chamber The range layout of the positioners is shown in Figures 1 and 2. As described previously, [1], the composite range was constructed from an x-y planar near-field scanner and a roll/azimuth spherical near field positioner. These axes were augmented by two additional axes. One was a variable slide axis above the azimuth and below the roll axis which permitted the throat dimension of the test positioner to be adjusted to accommodate antennas of various dimensions. This slide was further augmented by a standoff that served to allow the mounting of a planar array close up to the x-y scan plane when smaller probe antennas were used. Lastly there was a probe roll axis to accommodate single-ported probe antennas. The probe antennas were open-ended rectangular waveguide, both for planar and spherical near-field scanning. The anechoic chamber to accommodate this system was required to be as economical in floor space as possible. The installed chamber was 15 ft x 1 ft x 8 ft (H). The absorber selected to handle these frequencies was 5 and 1 inches thick. It was left unpainted so as to allow millimeter-wave performance. 2. Consideration of Positioning Error Millimeter-wave measurements in a near-field antenna measurement facility required significant attention to detail. Earlier approaches to the analysis of probe and AUT positioning errors [2] identify many potential sources of position uncertainty that must be controlled to make successful measurements. For near-field systems that operate at millimeter-wave frequencies, mechanical errors that affect the motion of the probe normal to the scan surface of the measurement become more significant than for lower frequency systems. This is because scanning of the ideal surface a plane, cylinder, or sphere, for example is perturbed by small imperfections in the mechanical apparatus.
Consider spherical scanning, for example. While scanning during data acquisition, the distance along the range axis between the probe tip and the center of rotation can vary due to angular bearing wobble, radial bearing runout, or the distortion effects of gravity. A mechanical system acceptable at lower frequencies may fall short at high frequencies because the undesired motion is a larger fraction of a wavelength. Undesired motion along the range axis contributes primarily to uncertainty in the measured RF phase. Phase uncertainty due to mechanical deviations can quickly become the limiting factor for accurate near-field measurements. The SNF positioner was analyzed by the methods outlined in [2], using a Quality Factor of 2 as a design goal. Careful analysis identified specific areas requiring tighter tolerances than conventional designs would have required, particularly for millimeter-wave operation. At 11 GHz, a Quality Factor of 2 implies a radius error limit of.54 inches, which drove the designers to significantly upgrade the class of bearing used in the azimuth axis. For Planar Near-Field measurements, a similar analysis of phase uncertainty due to planarity and other mechanical tolerances was also considered. In this case, other system accuracy requirements on planarity were already driving tolerances to a level that provided adequate phase uncertainty for all bands. Because of the minimum radius requirements for SNF operation and limited range of the mechanical slide, an optional horizontal standoff from the face of the roll axis was added to reduce truncation of the measurement surface during PNF operation. This is particularly advantageous when measuring thin antennas. Great care was taken during installation to achieve mechanical alignment consistent with design goals. The spherical positioner achieved axis intersection of roll and azimuth axes of.39 inches. Axis orthogonality was.15 degrees from normal. The x-y planar near-field scanner achieved positional accuracy of.2 inch, and a planarity of.16 inches RMS. To ensure realignment when reconfiguring between planar, cylindrical or spherical scanning, an autocollimation system was integrated. 3. The System Dynamic Range at Millimeter Band Referring to the System block diagram of Figure 2, the dynamic range of this measurement can be calculated. Given a WR-1 standard gain horn of nominal gain 24 i, an open-ended waveguide probe of nominal gain 7 i, a transmit power of +5 m, a mixer conversion loss 36, and frequency of 94 GHz, the power level of the received signal P A is calculated to be -73 m. The MI-1797 Microwave Receiver has a minimum detectable signal level of 11 m, so the expected dynamic range of this measurement system is (-73) - (-11) = +37 which agrees with our observations. Had we chosen to decrease the measurement distance by a factor of ten, say, to 4 inches, then the dynamic range would improve by 2 (2log(1) = 2 ) to +57, assuming far-field conditions. Even so, our results show the present dynamic range to be adequate for measuring the directivity of a +24 i pyramidal horn. The end result after transforming, was a far-field dynamic range better than 5 for a standard gain horn at 94 GHz. Please see Figures 7 and 8. 5. Results of Standard Gain Horn Measurements To check out and verify the performance of the range we employed pyramidal horns appropriate to each respective waveguide band. The pyramidal horns for X-Band through Ka-Band were designed as gain standards and had approximately 22 i of gain[3]. The pyramidal horn for W-Band was consistent with this design although not explicitly included in the original NRL layout [3] of gain horns. Extensive comparison of planar, cylindrical and spherical near-field measurements was made for the X-band of frequencies. The results of pattern comparisons were similar to those obtained earlier. The focus of testing for Ka-band was on planar near-field and spherical near-field results. We emphasize here the Ka-band and W-band results made with spherical nearfield scanning. We were especially interested in the Ka-band tests to check for rapid multi-frequency data acquisition speed. To test for this, we acquired a 21-frequency data set on a Ka-band standard gain horn, using spherical near-field scanning over 4π steradians. This data could be processed for the directivity at each frequency. It is well known that spherical near-field scanning yields directivity very accurately [4]. Thus we were very interested to see how our directivity versus frequency compared to the standard NRL curve. The result is plotted in Figure 4. Agreement with the MI Technologies published gain curve to within ±.25 and with an independently computed aperture integration directivity result [5] to within ±.15 was found. An example of a Ka-band pattern measurement result from the system is shown in Figure 5 & 6 for 3 GHz.
The contour plot exhibits 8 of dynamic range in 5 steps. The patterns are displayed over 6 of dynamic range. An example of a W-band pattern measurement result from the system is shown in Figure 7 & 8 for 94 GHz. The contour plot exhibits 8 of dynamic range in 5 steps. The patterns are displayed over 6 of dynamic range. One can observe a difference in the signal-to-noise ratio of the far-field patterns. The presence of noise is clearly more apparent in the 94 GHz contour plot than in the 3 GHz plot. We have found that to obtain accurate numbers for the directivity at W-band, we must integrate extensively to expand the signal-to-noise ratio. Making 21-fold multiple frequency measurements works well at Ka-band. At W-band more averaging is required to achieve the same dynamic range as for Ka-Band. We have been successful in making 3-fold multiple frequency pattern results at W- band. 6. Summary We have designed and fielded a composite near-field range capable of operation from 8.2 GHz through 11 GHz. We have demonstrated the performance by exhibiting pattern results for pyramidal horns. A far-field dynamic range greater than 8 at Ka-Band and greater than 5 at W-Band has been demonstrated. Very good agreement in radiation patterns between theory and measurement has been shown both for Ka- Band and for W-Band. 7. Acknowledgements The authors acknowledge our use of and the sharing of a MathCad application program written by Mr Keith Dishman that calculates the gain of pyramidal horns by use of textbook formulas, believed to be equivalent to the theory underlying the development of the NRL gain formulas. 8. References [1] Hess, D.W., Readily made comparison among the three near-field measurement geometries using a composite near-field range, pp., Proceedings, 23 AMTA Symposium, Irvine, CA. [2] Hess, D.W., "An expanded approach to spherical near-field uncertainty analysis," pp. 495-5, Proceedings, 22 AMTA Symposium, Cleveland, OH. [3] Slayton, W.T., Design and calibration of microwave antenna gain standards, NRL Formal report 4433, 9 Nov., 1954. [4] Hansen, J.E., Editor, Spherical Near-Field Antenna Measurements, pp.27-214, Peter Peregrinus Ltd., London, U.K. 1988. [5] Balanis, C.A., Antenna Theory, Section 13.4, Pyramidal Horns, John Wiley and Sons, Inc., New York, N.Y. 1997 Figure 1. Side Elevation View of Range and Chamber
X6 P1 C G1 R=43 inch G2 P2= P1 G1 G2 { } 2 λ 4 π R P REF P A Receiver MI-1797 Figure 2. Millimeter-wave RF Schematic -- Block Diagram -2-3 -4 8 9 1 11 GHz Signal Channel Reference Channel Figure 3. Example of Power versus Frequency at W-Band File is Output_AzElmergACQD-SNF- // Date is 7/13/24 6:45:3 PM 25.1 25. 24.9 24.8 24.7 24.6 24.5 24.4 24.3 24.2 24.1 24. 23.9 23.8 3 31 32 33 34 35 36 37 38 39 4 Frequency (GHz) Theta ':9. ':.(File: Output_AzElmergACQD-SNF-,7/13/24 6:45:3 PM) Theta ':9. ':-.25(File: Output_AzElmergACQD-SNF-,7/13/24 6:45:3 PM) Theta ':89.75 ':.(File: Output_AzElmergACQD-SNF-,7/13/24 6:45:3 PM) Theta ':89.75 ':-.25(File: Output_AzElmergACQD-SNF-,7/13/24 6:45:3 PM) (Disabled axis):18. (Disabled axis):.(file: Dishman_Gain_Standard_Template_,4/7/24 1:23:5 AM) (Disabled axis):18. (Disabled axis):.(file: Gain_Standard_Template_,4/7/24 1:23:5 AM) Figure 4.Measured and Computed Directivity, and Standard Gain versus Frequency
Figure 5. 2π SR Far Field Derived from Spherical Near-field Measurememts at 3 GHz File is Output_El_Cut_3GHz_mergACQD-SNF- // Date is 7/14/24 1:31:45 AM File is Output_Az_Cut_3GHz_mergACQD-SNF- // Date is 7/14/24 11:17:55 AM -1-1 -2-2 -3-3 -4-4 -5-5 -6 1 2 3 4 5 6 7 8 9 1 11 12 13 14 15 16 17 18 Theta ' -6 1-9 7 9-8 -7-6 -5-4 -3-2 -1 2 3 4 5 6 8 ':. Freq: 3. GHz Theta ':9. Freq: 3. GHz File is Output_El_Cut_3GHz_mergACQD-SNF- // Date is 7/14/24 1:31:45 AM File is Output_Az_Cut_3GHz_mergACQD-SNF- // Date is 7/14/24 11:17:55 AM -1-1 -2-2 -3-3 -4-4 -5-5 -6 9 1 11 12 13 14 15 16 Theta 17 18 ':. Freq: 3. GHz(File: Output_El_Cut_3GHz_mergACQD-SNF-,7/14/24 1:31:45 AM) Azimuth:. Freq: 3. GHz(File: Output_El_Cut_DishmanModel,7/12/24 6:4:42 PM) -6 1 2 3 4 5 6 7 8 9 Theta ':9. Freq: 3. GHz(File: Output_Az_Cut_3GHz_mergACQD-SNF-,7/14/24 11:17:55 AM) ElBar:9. Freq: 3. GHz(File: Output_Az_Cut_DishmanModel,7/12/24 5:23:26 PM) Figure 6. Pyramidal Standard Gain Horn Pattern at 3 GHz and Comparison to Theory
Figure 7. 2π SR Far Field Derived from Spherical Near-field Measurememts at 94 GHz -1-1 -2-2 -3-3 -4-4 -5-5 -6 1 2 3 4 5 6 7 9 1 11 12 13 14 15 16 17 18 8 Theta ' -6-9 -8-7 -6-5 -4-3 -2-1 1 2 3 4 5 6 7 8 Theta ':9. Freq: 94. GHz ':. Freq: 94. GHz -1-1 -2-2 -3-4 -4-5 -6-3 -5 9 1 11 12 13 14 Theta ':. Freq: 94. GHz(File: Output_ElCut_94GHz_MergAcd_RXAUT_1,7/14/24 2:34:31 PM) ':. Freq: 94. GHz(File: Output_ElCut_94GHz_DishmanModel_E_Plane,7/14/24 3:49:17 PM) 15 16 17 18-6 1 2 3 4 5 6 7 8 9 Theta ':9. Freq: 94. GHz(File: Output_AzCut_94GHz_MergAcd_TXAUT_,7/14/24 3:46:52 PM) Theta ':9. Freq: 94. GHz(File: Output_AzCut_94GHz_DishmanModel_H_Plane,7/14/24 3:46:52 PM) Figure 8. Pyramidal Standard Gain Horn Pattern at 94 GHz and Comparison to Theory 9