15 Dec 2016

RF & microwave power measurements: making the right field test choices

Giovanni D’Amore of Keysight Technologies looks at thermistor, thermocouple, diode & spectrum-analyser power measurement & the right choice for your application.

The output power of an RF or microwave system is a key determinant of its performance. The two key instruments for making measurements of RF or microwave power in the field are power sensors and spectrum analysers. Each approach has its strengths, and so making a choice of which to use involves trading off between accuracy, frequency range, dynamic range, portability, durability and warm-up time.

Power sensor architecture

A typical power sensor converts an RF/microwave signal to a DC or low-frequency analogue voltage waveform of around 100 nV, which is then amplified and filtered. The first trade-off here is to decide whether to reduce the filter bandwidth to improve measurement sensitivity, or to increase it to improve measurement speed. Whichever choice is made, the resultant signal is then digitised using an analogue-to-digital converter (ADC), so a microprocessor can handle further filtering and time averaging of the waveform.

Block diagram of two approaches to power sensor and power meter architecture

Figure 1: Block diagram of two approaches to power sensor and power meter architecture (Source: Keysight Technologies)

Different forms of power sensors handle signal processing in different ways. A USB-based power sensor handles its own signal processing (as shown in the black box in Figure 1). If the power sensor has a separate power meter, signal processing is handled in the meter (the red box in Figure 1). Some peak power sensors have additional signal paths to process average and peak measurements separately.

Power sensors usually have the highest measurement accuracy but require zeroing and user calibration to correct for frequency response, temperature drift and sensor aging. Keysight power sensors and meters have a 50 MHz reference oscillator for this purpose, whose output power is controlled to ±0.4%.

Thermal power sensors

RF/microwave power sensing is usually done with thermal or diode-based detectors.

Thermal sensors respond to the total power in the signal and report its true average power, regardless of modulation, by using the energy of the RF/microwave signal to change one of their electrical properties.

Sensor elements are usually a thermistor or thermocouple.

Thermistors warm up when subject to an RF input, so their resistance drops in a way that can be accurately measured in a bridge circuit. Drawbacks of thermistors include low RF measurement sensitivity (down to approximately -20 dBm), slow operation, sensitivity to ambient temperature, and the need to be connected to a power meter.

In thermocouples, temperature changes generate a voltage change that can be directly measured. Thermocouples are more rugged and less sensitive to ambient temperature than thermistors, making them more useful for field measurements. Their sensitivity is slightly greater, at approximately -35 dBm. Drawbacks include the need for regular calibration for temperature drift and sensor aging, using a reference oscillator

Diode-based power sensors

Diode-based sensors have a wider dynamic range than thermal-based sensors, but they have non-linear characteristics that need careful compensation.

Diode-based sensors rectify and filter the input RF/microwave signal using a diode and capacitor, as in Figure 2.

Diode rectification and filtering of an RF input

Figure 2: Diode rectification and filtering of an RF input (Source: Keysight Technologies)

The detected output voltage of a diode-based sensor is proportional to the square of the input voltage and therefore linearly related to the input power, within a region from -70 dBm to -20 dBm of input signal. Above 0 dBm, the diode characteristic shifts from a power-to-voltage to linear voltage-to-voltage relationship, through a transition region from –20 dBm to 0 dBm.

Drawbacks of diode-based power sensors include the need for access to a reference oscillator for calibration, a warm-up time of up to 30 minutes before use for greatest accuracy, and sensitivity to ESD and mechanical shock.

Putting diode-based power sensors to work

Here are three ways a diode-based power sensor can be used.

Measuring the average power of CW signals

With the right signal-processing and compensation techniques, the dynamic range of a diode-based sensor can be more than 90 dB. This is often done by saving correction factors in non-volatile memory to compensate for variations in input power, temperature and frequency.

Measurements in the higher-power transition and linear regions of a diode-based sensor are only accurate for continuous wave (CW) signals. This means they can be used to test frequency sources and the output power from amplifiers. Some diode-based sensors, such as the Keysight E4412/13A, are optimised for CW measurements, for which they have the lowest sensitivity and highest dynamic range of all the sensor types.

Measuring the average power of modulated signals

The correction factors used in the transition and linear region of a diode-based sensor are not accurate enough for the sensor to measure the average power of pulsed and digitally modulated waveforms.

It is possible to attenuate a high-power signal until it can be measured in the square-law region of the sensor. This reduces the sensor’s sensitivity, dynamic range and accuracy. Another approach is to use a sensor with two measurement paths, which can keep the diode operating in its square-law region at all power levels. These wide dynamic-range sensors, such as the Keysight E9300 E-Series, have one measurement path covering -60 to -10 dBm and a second path, with built-in attenuator, to cover -10 to +20 dBm.

**Measuring the peak power of modulated signals **

Measuring the peak power of signals with pulsed and modulated signals requires two levels of signal detection. The modulated signal first has to be detected using the diode sensor and converted to a wideband signal representing the time-varying envelope of the modulated waveform. The second detection occurs when the wideband signal is sampled and processed. Amplitude corrections can then be applied to the samples by adjusting the measured power to match the square-law, transition or linear regions of the diode sensor’s operation.

Tuned receiver power measurements

A spectrum analyser can also be used to measure average power. Handheld instruments such as the Keysight FieldFox can take this measurement with an accuracy that is typically ±0.5 dB across its frequency and temperature range.

The architecture of the tuned receiver used in spectrum analysers

Figure 3: The architecture of the tuned receiver used in spectrum analysers (Source: Keysight Technologies)

The front end of the circuit shown in Figure 3 has a down-conversion block using a mixer and local oscillator, plus a bandpass filter before the amplitude detector, to convert the RF input signal to a frequency at which it can be more easily filtered and detected. By sweeping the local oscillator across a frequency range, a spectrum analyser can display the power of a signal as a function of frequency.

The ADC samples the down-converted intermediate frequency (IF) for filtering and detection in the digital domain, in this case in two stages. In the first, there’s a bandpass filter with an adjustable bandwidth, which is referred to as its channel or resolution bandwidth (RBW). The next stage is for the signal’s amplitude to be detected and low-pass filtered, using a firmware algorithm that simulates an ideal square-law detector.

One of the advantages of spectrum analysers over power meters is their frequency selectivity, the ability to measure the power in a set RBW. This enables spectrum analysers to measure less powerful signals than power meters, by selecting a narrow RBW. In practice, a power meter may measure down to -70 dBm, while a spectrum analyser like the FieldFox can measure down to -154 dBm/Hz with a dynamic range greater than 105 dB.

RF power meters and spectrum analysers each have advantages and disadvantages. Understanding these should help users choose the approach that will bring them the accuracy, repeatability and traceability they need in their RF and microwave signal power measurements.

Page 1 of 1


About the author

Giovanni D’Amore is the Marketing Brand Manager at Keysight Technologies responsible for the EMEA and Indian RF & Microwave product line and market. Prior to his current role, Giovanni held numerous marketing, business development and applications support positions across several product lines with Hewlett-Packard, Agilent Technologies, and now Keysight Technologies. A regular presenter at conferences around the world and the author of several articles on microwave measurement techniques, Giovanni received a Masters degree in Electronic Engineering specialising in Microwave and Telecommunications from Palermo University (Italy).

Keysight Technologies Inc. is the world’s leading electronic measurement company, transforming today’s measurement experience through innovations in wireless, modular, and software solutions. With its HP and Agilent legacy, Keysight delivers solutions in wireless communications, aerospace and defence and semiconductor markets with world-class platforms, software and consistent measurement science. Keysight’s singular focus on measurement helps scientists, researchers and engineers address their toughest challenges with precision and confidence. With the help of their products and services, they are better able to deliver the breakthroughs that make a measurable difference. For more information visit Keysight Technologies.

Most popular articles in Test & measurement

  • DDR Memory Characterization Using a Mixed Signal Oscilloscope
  • RF Generator Experience for Medical Applications
  • RF & microwave power measurements: making the right field test choices
  • Realtime Spectrum Analyzers - what are they?
  • Current Sense Transformer Performance
  • Share this page


    Want more like this? Register for our newsletter








    A flexible approach to better antennas Nick Robins | Alpha Micro Components
    A flexible approach to better antennas
    For the Internet of Things to become as ubiquitous as forecasts suggests, a lot of Things are going to need Internet connections, many of them wireless.
    Training
    Online - Transmission Lines, S-Parameters & Smith Chart
    Understand these essential concepts without complex mathematics

    More training courses

    Whitepapers
    mmWave The Battle of the Bands
    5G mobile communications will use mmWave for many short range wide band links - but which microwave band will be used? Find out more in this informative white paper.

    More whitepapers










    Radio-Electronics.com is operated and owned by Adrio Communications Ltd and edited by Ian Poole. All information is © Adrio Communications Ltd and may not be copied except for individual personal use. This includes copying material in whatever form into website pages. While every effort is made to ensure the accuracy of the information on Radio-Electronics.com, no liability is accepted for any consequences of using it. This site uses cookies. By using this site, these terms including the use of cookies are accepted. More explanation can be found in our Privacy Policy