A wide range of Radio Frequency (RF) assemblies require testing during their development and production. These include amplifiers, mixers, filters, up converters, down converters, antennas, phase shifters, power dividers, power combiners, Integrated Master Assemblies (IMAs), Weapons Replaceable Assemblies (WRAs), Line Replaceable Units (LRUs) and radio subassemblies. Though there are many differences between the RF assemblies, they all require an
automatic test system that is accurate, reliable and traceable while simultaneously minimizing test times. Other important attributes of
instrumentation and automatic
test equipment (ATE) for RF assemblies are that it must be calibrated, fully-characterized and incorporate self-test functionality. The test platform must also include a digital control solution for triggering, communicating with Units Under Test (UUT) and interfacing with handlers, fixtures, etc. Finally, the software for the test equipment must also be considered. Typical software functionality would include sequencing, reporting, testing, calibration, and self-test. While it is certainly possible to develop a test system specific to each type of RF assembly; there are many benefits to developing a common RF test platform that includes the functionality listed above.
Keywords-Radio Frequency, RF, Automated Test Equipment, common platform.
I. INTRODUCTION
At its fundamental level of abstraction, a common RF test platform comprises four basic subsystems:
1. RF Instruments (Stimulus and Measurement)
2. RF Switching
3. Digital and Analog Control
4. Software.
This paper discusses the high-level design and measurement uncertainty analysis of a common, or general-purpose, RF test platform. The design followed the customary process of generating requirements, generating options for satisfying those requirements and evaluating those options based on performance criteria. For this design the three major performance criteria were maximizing measurement accuracy, minimizing test time and minimizing total cost-of-ownership. The remainder of this paper discusses the options we considered for each subsystem in a general context and then presents the options we chose and ultimately implemented. Finally, the paper concludes with a measurement uncertainty analysis of the platform we developed. The measurement uncertainty results the paper presents are specific to the system we developed, but the analysis is presented in a way such that it could be applied to other RF test
systems
II. RF INSTRUMENTS
Not too many years ago there was little choice for the designer of an RF test system as to instrumentation architecture. There were different vendors, of course, but it ultimately was a choice between rack-mounted, bench-top
instrumentation including oscilloscopes, signal generators, network analyzers and signal analyzers.. In a traditional architecture these instruments are multiplexed to the UUT via an RF switch matrix, often called an RF Interface Unit (RFIU).
The choices in architecture are changing with the growing popularity of synthetic instruments. Synthetic instruments use combinations of smaller modules to implement the functionality of larger bench-top instruments. For example,
a vector signal analyzer could be composed of three modules: a signal generator, a downconverter, and a digitizer.
PXI, PXIe and VXI are all common platforms for synthetic instruments.
Proponents of synthetic instruments cite advantages in terms of increased system versatility, decreased threat of obsolescence, greater ease of calibration and lower cost (see for example “Synthetic instruments in automatic test systems” by John Stratton in the November 2004 edition of Defense Electronics, pages 20 – 23). While there are points to be made in favor of synthetic
instrumentation in these regards, the proponents of bench top instruments present many counter points and other criteria that favor a traditional architecture. Historically the easiest point to make in favor of bench top instruments has been higher frequency range, but the newer generations of synthetic instruments are making great advances in frequency range. Spectrum and vector analysis to 26.5 GHz and beyond is currently possible with synthetic instruments.
The makers of traditional-architecture instruments have of course noticed the competition from synthetic instruments and are adding to the feature-sets of bench-top instruments to incorporate some of the advantages of synthetic instruments. Ultimately for our design we chose a traditional
instrumentation architecture. We worked with a number of vendors of synthetic instruments and researched their offerings, but could not find anything currently available that would meet our frequency range and measurement accuracy requirements. Following is the list of RF instruments we chose for our common RF test platform.
• Agilent N5244A PNA-X Network Analyzer
• Agilent N9030A PXA Signal Analyzer
• Agilent E8267D Vector Signal Generator
• Agilent N8241A Arbitrary Waveform Generator
• Agilent E4419B Power Meter
• Agilent 346CK01 Noise Source
• Agilent N4692A E-CAL module
Based on our analysis, these instruments give us the greatest RF accuracy with the shortest test times and the lowest overall cost-of-ownership.
→
|
