Distributed RF Systems for Antenna Measurements
By Dave Fooshe and Bert Schluper, Nearfield Systems Inc.

It is well known that modern, high-performance antenna measurements require fast sources and receivers. What’s often overlooked, however, is the role that instrument and component location within the antenna range play in determining system performance. Locating RF sources and receivers close to the transmitting and receiving antennas or using remote mixers, amplifiers and multipliers allow the use of shorter or lower-frequency cables, offering the potential to dramatically increase the available power level at the transmitting antenna. Use of fiber optics is also becoming an option for transmitting RF signals in distributed RF systems. Automated configuration control of the distributed range may also be achieved using remotely controlled RF switches.

Nearfield Systems Inc. (NSI) has a long history of implementing distributed RF systems on large antenna ranges. As an example, the 33m x 16m vertical near-field scanner installed at Toshiba in 1998 was then considered the largest vertical near-field range in the world [1]. With cable lengths of 60m distance from the control room to the scanner x-carriage and 26m additional distance to the probe carriage, the system represents a unique implementation of a distributed RF system for antenna measurements. The Toshiba 50 GHz RF system includes remote mixers, control of remote RF sources, remote PIN switches and a remote frequency multiplier/amplifier unit.

At the other end of the spectrum is the small, low frequency range that simply uses a vector network analyzer or integrated frequency converter in a central location or standalone configuration.

The type of RF system best suited for a particular antenna range depends upon a number of factors, including the measurement technique, frequency range and antenna parameters of interest. The choice of measurement technique, i.e. planar, cylindrical or spherical near-field, far-field or compact range, directly impacts the performance of the system by influencing the range dimensions and, therefore, the RF cable lengths. Long RF cables increase the loss budget and detract from the ability of the system to measure low side-lobe levels. Similarly, the electrical requirements may influence the measurement technique chosen. The inter-dependencies involved in specifying an antenna measurement system highlight the need for a systematic approach to identifying and analyzing system requirements before settling on a particular measurement technique or a specific RF system configuration [2].

The goal of this article is to demonstrate the performance improvements that may be expected by implementing a distributed RF system. For example, how large should the antenna range be before one begins to consider a distributed RF system? A standalone system may work well up to 18 GHz for a 12 ft. x 12 ft. near-field scanner, but it could operate to 40 GHz with a 6 ft. x 6 ft. scanner. How does this inverse relationship extend to spherical, far-field or combination ranges? Since the size of the range also determines the dynamic range of the measurement system, the antenna test requirements are also an important factor.

This article will compare three different antenna ranges and show the differences in performance depending upon the choice of configuration. Instrumentation control, range automation and the use of fiber optics for RF transmission will also be discussed.

RF System Assumptions
In order to show the advantages of the distributed RF system approach, simplifying assumptions have been made in order to present three different antenna range examples for comparison. The assumptions include:

• A 12 ft. x 12 ft. planar near-field scanner located inside a 9m x 5m x 6m chamber was selected as an average size, typical of many existing ranges.

• A frequency range of 1 to 40 GHz was chosen, representative of many of today’s antenna ranges.

• Cable length distance of 12m from RF rack to the Antenna Under Test (AUT). Cable length distance of 20m from RF rack to probe.

• Probe gain is 6 dBi, AUT gain is 30 dBi.

The following sections will show three different RF configurations while keeping the above parameters constant. This will provide a means of comparing the performance improvement that may be expected with a distributed RF system and also by using a remote frequency multiplier.

Configuration 1 - Standalone RF
The first configuration, a standalone RF system, consists of a 40 GHz Agilent Performance Network Analyzer (PNA) located in a control room adjacent to the antenna range. The standalone system consists of 40 GHz RF cables from the PNA to the AUT and also from the probe back to the PNA. Figure 1 shows a simplified block diagram of the standalone RF configuration.

Configuration 2 - Remote Mixers
The remote mixer RF system uses a Distributed Frequency Converter (DFC) with mixers located at the probe and near the AUT. In the AUT Transmit (AUT Tx) configuration, the RF cable runs from the frequency source (PNA) to the AUT, but the mixer on the receiving end is located near the probe, thereby reducing the total RF cable length requirement. Figure 2 shows the addition of the DFC and remote mixers. The 40 GHz cable runs from the PNA to the coupler, then to the AUT. A reference mixer is located near the AUT. A test mixer is located near the probe, thereby reducing the 40 GHz probe path from 20m to 1m.

The NSI-RF-5942 Distributed Frequency Converter, shown in Figure 3, includes two remote mixers and an LO/IF distribution unit and supports long LO cables with up to 30 dB of loss. This allows LO cable lengths of over 30m while still using fundamental mixing to 18 GHz.

Configuration 3 - Remote Multipliers
The third example, shown in Figure 4, enhances the remote mixer RF system by locating a frequency multiplier near the AUT. This reduces the frequency of the RF cable from the source to the AUT to 20 GHz instead of the more expensive and higher loss 40 GHz cable, while also allowing the use of the lower cost 20 GHz PNA. The NSI-RF-5994 provides an amplifier, multiplier and coupler near the AUT. The amplifier alone may be used below 20 GHz when the multiplier is not required. The amplifier and multiplier are located prior to the coupler, resulting in cancellation of noise in the S21 signal due to the common test and reference path.

Comparison of Configurations
A comparison of the power budget factors for the three configurations is shown in Table 1.
The standalone system power losses vs frequency are shown in the bottom (red) plot of Figure 5. Due primarily to cable loss and AUT/probe loss, the power at the input to the PNA is -39 dBm at 1 GHz and decreases with frequency to below -90 dBm above 25 GHz. This would not be considered an acceptable level of performance and would likely result in the rearrangement of equipment to shorten the cables. An external amplifier could be added, however its noise figure would reduce the RF system sensitivity, since it could not be placed in a position common to both test and reference paths.

The remote mixer system power losses are shown in the middle (green) plot of Figure 5. The power at the input to the PNA is -30 dBm at 1 GHz and decreases to -70 dBm at 40 GHz. This represents an improvement over the standalone case, but performance suffers at higher frequencies.

The remote multiplier system power losses are shown in the top (blue) plot of Figure 5. The power at the input to the PNA is -11 dBm at 1 GHz and decreases to -24 dBm at 40 GHz. Although manual reconfiguration is required going above or below 20 GHz, this represents a significant improvement over the either of the prior cases. In this case, attenuation may even be required at lower frequencies to avoid mixer or receiver saturation.

Remote Equipment, Fiber Optics and Range Automation
Additional performance can often be gained by remotely locating an RF source or network analyzer near the probe or AUT. General Purpose Interface Bus (GPIB) extenders have been used for this purpose, however, many of today’s instruments are equipped with a Local Area Network (LAN) interface, which provides a simple, low cost interface to the data acquisition computer and the ability to control instruments over much longer distances. The NSI Panther 9000 receiver is a LAN instrument and offers support for multiple high speed beam controllers, which may be an attractive option for timing and sequence control of distributed instrumentation [3].

Fiber optic links have been used on the antenna range to control instrumentation, distribute time base, LO and RF signals [4]. The low loss and extremely wide bandwidth of fiber optic cable make it an attractive alternative to coax for the transmission of LO and RF signals. Fiber optic links are now available that are capable of transmitting RF signals in the 0.1 GHz to 18 GHz frequency range, making them broadly applicable to antenna range applications. NSI is currently developing a 700m fiber optic link to transmit the RF signal received from the AUT on a far-field range back to the test mixer in the control room. The fiber optic link contributes approximately 1 dB of additional system noise while providing additional advantages such as low loss, high isolation and fundamental mixing to 18 GHz.

Distributed RF systems often require manual reconfiguration of remote cables and components, resulting in configuration control and test repeatability challenges for antenna range operators. NSI has developed the NSI-RF-TRU Range Transmit Receive Unit (TRU) to solve this problem; see Figure 6. The TRU includes a LAN interface to easily control remote RF modules with built-in switches, mixers, amplifiers, attenuators and other components. The TRU allows the range operator, or an automated script, to easily reconfigure a very complex distributed RF system in a matter of seconds [5].

Summary
Distributed RF systems are a necessity for many large high frequency antenna ranges, but any range can benefit from the performance analysis and tradeoff to determine if a distributed RF system would improve performance or measurement efficiency. As shown with these three range examples, moving to a distributed configuration can have a dramatic impact on RF system performance.

Fiber optic solutions, remote instrumentation and automated configuration control are other options currently available for antenna RF systems.

References
[1] T. Speicher, S. Sapmaz, M. Niwata, “33m x 16m Near-field Measurement System,” AMTA Proceedings 1998.
[2] D. S. Fooshe, M.J. Schultz, “How to Specify an RF System for Antenna Measurements,” AMTA Proceedings 2000.
[3] D. Slater, “Next Generation Phase Coherent Instrumentation Receiver,” AMTA Proceedings 2007, p. 302, St. Louis, MO, Nov 2007.
[4] G.J. Matyas, “High Speed Fiber Optic Remote Receiver Link for Improved Antenna Measurements,” AMTA Proceedings 1992, p. 14-4.
[5] D.S. Fooshe, “Improving Automation for Antenna Ranges,” AMTA Proceedings 2006, p. 339, Austin, TX, Oct 2006.

Nearfield Systems Inc.
www.nearfield.com
Email this article to a friend!