A HEMI-SPHERICAL NEAR-FIELD SYSTEM FOR AUTOMOTIVE ANTENNA TESTING

 

Pieter N. Betjes, Dieter Pototzki* & Daniλl Janse van Rensburg
Nearfield Systems, Inc
19730 Magellan Drive, Torrance, CA 90502-1104, USA
E-Mail: pbetjes@nearfield.com & drensburg@nearfield.com
Worldwide web: www.nearfield.com

*Nippon Antenna Co., Ltd. (EU) R&D Center
Fraunhoferstr. 3, D-25524 Itzehoe, Germany
E-Mail:adp@nippon-antenna.com

 

ABSTRACT

A hemi-spherical near-field test system with added far-field capability is described. The facility has been constructed for the characterization of automotive antennas. The test system consists of an 11m tall dielectric gantry, a 6.5m diameter in-ground turntable and a 28m-diameter radome enclosure. Special software required to compensate for the reflectivity in the facility and the hemi-spherical truncation was developed and forms an integral part of this test system. The characteristics of this facility are described in this paper and measured data is presented.

Keywords: Vehicular antennas, Spherical near-field, Automotive testing.

1. Introduction

The sophistication of RF sub-systems being integrated in modern automobiles is leading to an increased demand for test systems allowing in-situ antenna testing. Unwanted interaction between RF sub-systems requires that antenna developers be able to do these tests as early in the development cycle as possible. Test times also have to be minimized and confidentiality of test articles ensured.

Since many of the antennas in question are electrically small it is often found that the entire body of the vehicle is excited during radiation and therefore needs to be considered part of the measurement. This drives the need for a large turntable and physical sizing of the test area.

Far-field ranges have traditionally been selected for this type of testing since it is a simple and quick process. However, the requirement for high elevation angle coverage (satellite based receiver systems) prescribes a technique where hemi-spherical antenna pattern testing is performed. A spherical near-field test system offers a unique solution in principle. However, mounting of a very heavy object under test is impractical for the traditional phi/theta configuration and an alternative solution like that depicted in Figures 1 & 2 has to be considered. This type of test system offers a practical solution to the test problem in that combined motion of a probe antenna and the object under test, allows for spherical data acquisition covering one half of the spherical surface. The configuration also allows integration of a conducting ground plane as well as a radome enclosure for weather protection and confidentiality.

The characteristics of this newly developed facility are described in the following section of this paper.

Figure 1: Layout and coordinate systems of measurement facility: View from far-field tower.

Figure 2: Layout and coordinate systems of measurement facility: Side view (far-field tower distance not to scale).

 

2. Facility Description

A facility as depicted in Figures 1 & 2 has been constructed by NSI, The Howland Company, EMV GmbH and Nippon Antenna in Itzehoe, Germany. It operates as a combined far-field and spherical near-field test range and can be used for antenna testing from 50 MHz to 6 GHz.

The facility contains a 6.5m diameter turntable with a weight capacity of 4500 kg, shown in Figure 3. (This turntable represents the phi-axis of the spherical near-field test system.)

Figure 3: Picture of the in-ground 6.5m diameter turntable.

A probe antenna is mounted on an 11m tall gantry (visible in Figure 5), which allows for elevation angle acquisition of 0 – 90°. (This gantry represents the theta-axis of the spherical near-field test system.) The probe is capable of acquiring two polarizations. The gantry structure is dielectric to minimize electrical interaction with the antenna under test. The test system is enclosed by a 28m-diameter radome and the turntable surface as well as the facility floor is metallized. The radome is shown in Figure 4 and the metallized interior ground plane is shown in Figure 5.

Figure 4: Exterior view of 28m-diameter radome.

Figure 5: Conducting ground plane, AUT and dielectric gantry shown in vertical position.

A tower with far-field source antenna is located at 120m from the radome and the source antenna is height-adjustable and contains a polarization positioner as shown in Figure 6. The ground plane between the radome and the tower is also metallized (as shown in Figure 7) to implement a highly repeatable ground reflection range.

Figure 6: Far-field tower source antenna and polarization positioner.

Figure 7: Far-field tower and conducting ground plane extending to test area.

This system configuration allows for single axis far-field measurements using the far-field source tower & turntable and hemi-spherical far-field measurements using the gantry & turntable. The latter case is a valid option only when the 11m-gantry height meets the far-field criteria. Hemi-spherical near-field measurements are possible using the gantry & turntable for all cases where there is no reactive near-field coupling of the antenna under test and the probe.

A graphical display of the frequency regions vs. size of the object under test (AUT) is given in Figure 8.

Figure 8: AUT dimensions and frequency regions for which the different measurement modes apply.

This graph depicts that for a fixed antenna of 1m maximum aperture dimension, the gantry & turntable can be used for hemi-spherical far-field testing from roughly 100 MHz up to a frequency of 1 GHz. For lower frequency testing the far-field tower should be used and for higher frequencies, spherical near-field testing is required.

The gantry can be lowered to below the level of the ground plane (indicated in Figure 2) and covered with steel panels, to allow undisturbed far-field measurements (using the outdoor far-field tower). This requires that the axis of the gantry is approx. 50 cm below ground level, so that the usable measurement range for the gantry is limited to approx. 87°.

The facility has a below ground control room and a preparation area adjacent to the control room with car elevator to allow measurement-preparations to be performed while RF testing is ongoing.

3. Mechanical Performance

The spherical near-field technique relies heavily on the accuracy of the acquisition process and system mechanical alignment [1 - 3]. Design of a large test system as described here requires an understanding of the sensitivity of various mechanical probe and AUT positioning inaccuracies. A theoretical study of some of these parameters was described in [4] and lead to a mechanical design specification.

Table 1 lists the travel and velocity requirements for the various axes of interest.

Axis

Travel Range

Travel Speed

Turntable

0° – 360°

12°/s

Gantry

0° – 90°

0.5°/s

NF Polarization

0° – 90°

N/A

Tower height

1 – 11m

N/A

Tower polarization

0° - 90°

N/A

Table 1: Travel limits and velocities for mechanical axes.

Of these axes, the ones associated with the hemi-spherical near-field acquisition process determine data accuracy. From the work described in [4] the design specifications for theses were derived and are shown in Table 2. Also shown in table 2 are the realized values for these two axes.

Parameter

Design figure

Realized figure

Turntable

+/-0.2°

0.002°

Gantry radial error

10mm

1mm

Gantry angle error

+/-0.25°

<+/-0.03° (Corrected)

Table 2: Design vs. realized turntable and gantry specifications.

4. RF Sub-system

The RF sub-system used in this facility is a Rohde and Schwarz ZVC vector network analyzer with external amplifier to provide appropriate power levels in the design frequency range of 50 MHz to 6 GHz. This RF sub-system is controlled by the NSI 2000 software and allows for multi-frequency testing on the fly.

The probe antennas in use (both for near-field and far-field measurements) are log-periodic array antennas, which provide broadband performance and eliminate the need for changing probes when covering wide frequency bands.

The NSI 2000 software also enables the time-domain gating capability of the analyzer, which can be used as a measure to reduce reflections during testing, should the AUT bandwidth allow for this.

5. Reflection Suppression Software

The radome enclosing this test facility is not an optimal environment for antenna testing due to reflections from the internal surface. To counter this effect a new algorithm was developed by NSI that suppresses the effect of these and other facility (i.e. gantry) reflections. This method relies on the spherical near-field conversion process and is described in more detail in [5]. An example of the performance of the algorithm is given in Figure 9 & 10.

Figure 9 depicts a radiation pattern cut for a standard gain horn taken in the radome enclosed facility (red) and the effect of the reflections is notable. Also shown in Figure 9 is the same radiation pattern cut taken in an anechoic chamber (blue).

Figure 10 depicts the same comparison but with the reflection suppression enabled and the improvement is evident.

Figure 9: Comparison of standard gain horn measurement in radome enclosed hemi-spherical system with measurement in anechoic chamber – Without reflection suppression.

Figure 10: Comparison of standard gain horn measurement in radome enclosed hemi-spherical system with measurement in anechoic chamber – With reflection suppression.

6. Hemispherical Boundary Treatment

The spherical near-field to far-field conversion algorithm expects measurement data on an entire spherical surface. The described facility uses hemi-spherical scanning and the omission of the lower hemi-sphere of data will introduce an error in the form of a far-field ripple. It is therefore required to address this abrupt data termination properly to reduce these induced errors.

A variety of boundary treatment schemes were evaluated and the best performance was found to be given by a combination of mirroring the data of the upper hemi-sphere into the lower hemi-sphere while also applying a transitional attenuation function. Figure 11 shows a comparison of a full sphere measurement (in an anechoic chamber), a truncated measurement without treatment and a truncated measurement with the aforementioned treatment. The red radiation pattern in Figure 11 is the full sphere reference pattern. The blue pattern represents the near-field truncated case without any treatment and the far-field induced ripple is evident. The green pattern represents the case where the boundary treatment has been applied and the reduction in the far-field ripple is evident.

Figure 11: Comparison of the far-field from full sphere data, truncated hemi-spherical data and hemi-spherical data treated with the described boundary treatment.

7. Summary

This paper describes a new hemi-spherical near-field & far-field combination system that was recently completed. The facility is enclosed in a radome, which provides weather protection and a confidential test environment. The facility consists of a conventional in-ground turntable and a newly developed dielectric gantry providing low levels of interaction with the AUT during testing.

Although the test approach is in principal well known, new software was developed to suppress unwanted reflections from the radome enclosure and this in itself represents a new development for automotive testing, making test data more reliable in an environment which is not as well behaved as traditional anechoic chambers.

The hemi-spherical nature of the acquisition system also required a new truncation treatment to be developed to reduce the errors introduced by this truncation. Results have been presented to demonstrate this aspect and work is continuing to improve upon this further.

8. References

[1] Newell, A. C., Hindman, G., "The alignment of a spherical near-field rotator using electrical measurements" In the proceedings of the 19th annual AMTA Meeting and Symposium, Boston, MA, 1997.

[2] Newell, A. C., Hindman, G., "Quantifying the effect of position errors in spherical near-field measurements", In the proceedings of the 20th annual AMTA Meeting & Symposium, Montreal, Canada, 1998.

[3] Newell, A.C., "The effect of measurement geometry on alignment errors in spherical near-field measurements", In the proceedings of the 21st annual AMTA Meeting & Symposium, Monterey, California, 1999.

[4] D. Janse van Rensburg, "Parametric study of angular and radial probe positioning errors in a large spherical near-field automotive antenna test system", In the proceedings of the ANTEM Symposium, August 2004, Ottawa, ON, Canada.

[5] Greg Hindman, Allen Newell, “Reflection Suppression in Large Spherical Near-Field Range”, In the proceedings of the 27th annual AMTA Meeting & Symposium, Newport, RI, 2005.