Herzlich Willkommen beim beam-Verlag in Marburg, dem Fachverlag für anspruchsvolle Elektronik-Literatur.


Wir freuen uns, Sie auf unserem ePaper-Kiosk begrüßen zu können.

Aufrufe
vor 5 Jahren

8-2018

  • Text
  • Komponenten
  • Technik
  • Radio
  • Filter
  • Oszillatoren
  • Quarze
  • Wirelessemv
  • Messtechnik
  • Bauelemente
  • Measurement
  • Frequency
  • Wireless
  • Simulation
  • Models
  • Devices
  • Filters
  • Messungen
Fachzeitschrift für Hochfrequenz- und Mikrowellentechnik

RF & Wireless High-Power

RF & Wireless High-Power RF Measurement: Techniques and Methods Table 1: Thermistor-Based Directional Coupler Assembly, Representative Uncertainty Analysis TEGAM INC. www.tegam.com AR Technologies www.arworld.us Demand for high-power RF and microwave device calibration is higher than ever. Technological advances in automation, semiconductor, and communications applications are largely responsible for this increased demand, and have pushed the traditional limits of high power RF metrology to higher frequencies and higher power levels. Frequency and power are not the only boundaries being challenged however. These new and improved technologies also bring a mandate for lower uncertainties. Low-uncertainty high power RF measurement is especially critical for industries such as semiconductor manufacture, but it’s important to all industries involved in high power applications in both commercial and governmental sectors. Low-uncertainty high-power RF metrology poses unique challenges and requires specialized equipment and measurement techniques to ensure repeatable, accurate measurements. In this paper, we will provide a broad overview of commonly-used high-power RF measurement techniques, and examine the advantages and disadvantages of each. We will discuss the equipment required, typical measurement uncertainties, and the power and frequency limits of each. Lastly, we will identify the applications best served by each methodology. See Table 2 for a summary of each of the methods. For purposes of this paper, measurement uncertainties have been normalized to 2-sigma to facilitate easy comparison between methodologies. However, be aware that uncertainties may be presented at different confidence levels between manufacturers, and sometimes inconsistently between product lines from the same manufacturer. Always carefully read the product datasheets and manuals to properly characterize reported uncertainties. General Considerations High-Power Measurement Equipment High-power RF Test stations typically require specialized, purpose-built equipment. Highquality signal generators, amplifiers, and filters are necessary to supply low-distortion test signals at the measured power and frequency points. Building such a test station can be an expensive proposition, so thoroughly understanding the options available is even more important. Filters are especially critical as minimal harmonic distortion is required to ensure accurate measurements. Low-uncertainty measurements often require 2nd harmonic distortion to be better than -50 dBc, which can be challenging for amplified CW signals. Filters must be designed for the particular frequencies and power levels to be tested, with minimal 56 hf-praxis 8/2018

RF & Wireless VSWR and insertion loss across the specified measurement range. Filters may be stand-alone, or designed into measurement devices. But with either configuration, it’s important to verify the system’s harmonic suppression characteristics prior to making traceable measurements. Test stations must also include a load capable of dissipating the expected power levels to be measured. Measurement Considerations High-power RF measurement can present challenges more easily handled in lower-power applications. When measuring high-power, measurement errors such as port reflection and mismatch, insertion loss, power linearity, and tolerance to thermal transients become significantly more important. These topics are beyond the scope of this paper, but will be discussed in detail in upcoming releases from TEGAM. Please visit www.tegam.com, or call us at 440-466-6100 for more information regarding high-power RF measurement instruments, uncertainties, and procedures. Techniques and Methods RF Wattmeter The most straight-forward method for measuring highpower RF is the use of a simple wattmeter. Available in a wide array of designs, wattmeters are a cost-effective method for RF measurement in the 100 mW to 10 kW power range. Wattmeters typically use a diodedetection circuit to measure RF power across a transmission line. The wattmeter is connected inline with the device under test, and can measure both forward and reflected power. These instruments are often designed to be portable with the indicating meter and measurement elements packaged as a single unit. The portability and versatility of wattmeters make them well-suited for field work. However, these conveniences come at the expense of accuracy. Wattmeters typically provide measurement uncertainties in the 3% to 10% range. This is often acceptable for field applications such as tuning an antenna for optimal impedance or making a “go, no-go” measurement of a radar station. For metrological measurements however, these high measurement uncertainties are unacceptable. Directional Power Sensors Directional power sensors are similar to traditional RF wattmeters. They typically employ a diode-detection circuit that converts RF power to a voltage, and can be used to measure both forward and reflected power. Some models provide the ability to change connector styles, thereby supporting a wider range of test devices. Similar to wattmeters, directional power sensors are designed to cover relatively narrow frequency bands necessitating the purchase of multiple models for broadband applications. However, there are several key differences as well. First, directional power sensors typically do not include a meter or other indicator on the sensor. Instead, they require an external meter. Secondly, directional power sensors often provide improved measurement uncertainty over traditional wattmeters. At specific calibrated frequencies and power levels, directional power sensors can be as good as ± 2%. Uncertainty can be considerably worse at points other than the calibration points though. Users should therefore carefully examine the sensor’s measurement uncertainty at all frequencies and power levels of interest before making traceable measurements. Directional power sensors provide an acceptable bridge between field and laboratory work. However, the requirement of a separate meter / indicator makes them less portable than RF wattmeters, and they have narrower frequency ranges and worse uncertainties than the directional couple assemblies and calorimeters discussed below. Directional Coupler Assemblies Directional coupler assemblies can be used similarly to directional power sensors, but rely on fundamentally different technologies to make power measurements. They utilize a low-power RF sensor paired with a directional coupler to measure RF power levels beyond the usual power range of the sensor. They also often include a large heat sink to dissipate heat generated at the coupled port of the directional coupler. Dissipated RF power is converted to heat energy inside the coupler, which can induce variability in the coupling coefficient. These variabilities are difficult to predict, and can significantly impact repeatability and absolute measurement accuracy. A heat sink properly positioned and sized to wick away thermal changes within the coupler will mitigate these uncertainties. Directional coupler assemblies can be configured using either diode- or thermistor-based sensors. The dynamic range of the assembly is primarily determined by the coupler used, and the assembly’s overall measurement uncertainty is the combination of the directional coupler power linearity and directivity, and the calibration factor uncertainty of the sensor. In short, the better the sensor and coupler, the better the final measurement uncertainty. Directional coupler assemblies provide two major advantages over the directional power sensor approach. First, coupler assemblies can cover a much wider frequency bands. Compared to a directional power sensor, coupler assemblies are relatively flat over wide frequency ranges and can be calibrated to provide a reasonably accurate, singledevice solution for wideband power measurement. Second, directional coupler assemblies typically have better measurement uncertainty across their frequency and power ranges than directional power sensors, with thermistor-based assemblies providing better uncertainties than diode-based assemblies. Diode sensors typically report power readings relative to a reference, usually 50 MHz. In practical terms this means that before each use, the sensor must be disconnected from the directional coupler assembly, connected to the 50 MHz reference port on the system power meter, standardized, then reconnected to the assembly. This sensor disconnect and reconnect introduces a significant and sometimes difficult-to-calculate uncertainty to the final measurement. Further, the 50 MHz reference itself will contribute a sizable uncertainty to the measurement. Thermistor-based assemblies however measure absolute power and can therefore be permanently fixed to the coupler assembly, resulting in muchimproved measurement uncertainty. This combined with the better uncertainty of thermistor sensors generally, results in a better overall uncertainty compared to diode-based assemblies, and much better than directional power sensors. Uncertainties of better than 1% can be achieved with a properly calibrated thermistor-based coupler assembly. Linearity is also an important consideration when selecting between diode- or thermistorbased coupler assemblies. Linearity measures the response of a sensor compared to changes to input power. For instance, if input power is doubled, a perfectly linear sensor’s measured power will also double. Diode sensors have a wide dynamic range, with some popular sensors covering ranges as large as -70 to +30 dBm. However, nonlinearity across this range can be as much as 5%, depending on the sensor design. When calculating total system uncertainty, it’s important to account for this nonlinearity. Thermistor sensors on the other hand have a limited dynamic range, but excellent linearity across that range. A typical ther- hf-praxis 8/2018 57

hf-praxis

PC & Industrie

© beam-Verlag Dipl.-Ing. Reinhard Birchel