Characterizing and Troubleshooting Digital RF Amplifier Systems
New wireless technologies can wreak havoc on RF PAs designed for voice-only communication.
With the emergence of high-speed data services on the wireless mobile networks, new challenges have been placed on the design and operation of power amplifiers. The bursted nature of new wireless access technologies (3GPP - HSPA, LTE, WiMax, and 3GPP2 - 1xEV-DO) can wreak havoc on the modern amplifier design that previously had been designed for voice-only communication.
The dynamic fluctuation of the asymmetric demands for delivering high-speed data services, coupled with a desire to reduce operations expense with improved efficiency, has lead to a new class of Digital RF amplifiers, the power of digital computing applied to RF amplifiers.
Digital RF amplifiers employ complex linearization techniques enabled by the advances in DSP and availability of low-cost CMOS. By taking advantage of digital corrections, the amplifier efficiency can be maximized by operating near saturation. These techniques often require a feedback loop to sense and correct non-linearities adaptively during all operating conditions. However, these new amplifiers must be developed to assure proper spectrum and modulation quality behavior during the dynamic operating conditions.
The linearization techniques with adaptive feedback loops can exhibit short-term transients when not properly implanted due to poor behavioral modeling, software errors, and memory effects. These transients can reduce or limit the overall communication link efficiency and represent a new class of challenges for the wireless service providers and network equipment manufacturers.
Due to the transient bursted nature of the new wireless signals and the desire to fine tune operating efficiency and spectrum performance at an acceptable level of modulation quality, it is imperative that correlated analysis of spectrum (ACPR), amplitude statistics (CCDF), and modulation quality (EVM) can be assessed on the same set of time-varying data.
Fortunately, advanced Real-Time Spectrum Analyzers (RTSAs) are available to facilitate the acquisition, measurement and analysis of digitally modulated signals. With wide capture bandwidth, deep memory and inherently correlated multi-domain measurements, these instruments enable the efficient and accurate characterization and troubleshooting of today’s Digital RF amplifier systems.
Figure 1: The RTSA acts as the receiver in a Tx/Rx pair when measuring vector signal parameters
Digitally Modulated Signals
When a RTSA is used to measure the vector parameters of modulated signals, the instrument will act as the receiver in a Transmit/Receive (Tx/Rx) pair. Figure 1 illustrates the components of a generic Tx/Rx chain and the role the RTSA plays in replacing the Rx function. The receive chain begins with a low-noise RF amplifier tuned to the receive frequency. In an RTSA, a pre-amplifier may be used for low level signal measurements, but is not needed for high level measurements of transmitters.
Like a receiver, the RTSA contains an intermediate frequency (IF) filter for spurious and interfering signal control; its bandwidth is that of the instrument’s capture bandwidth, which may allow unwanted signals into the measurement.
Vector measurements of digitally modulated signals require the incoming signal to be compared to an ideal signal of the same modulation type and data. The signal analyzer must be aware of, and capable of reproducing, the modulation parameters of the signal, including frequency, symbol rate, modulation type, transmit/receive filters and transmitted symbol values.
Once the signal has been demodulated and the reference signal constructed, vector measurements can be performed, such as Error Vector Magnitude (EVM), Magnitude Error, Phase Error, Origin Offset, Gain Imbalance and rho (see Figure 2).
Figure 2: Examples of vector measurements made by a RTSA or VSA. Other panels display
magnitude vs. time, EVM vs. time and constellation display of the same time period.
PAR and CCDF
Modern amplifiers use sophisticated techniques to limit the Peak-to-Average-Ratio (PAR) of the amplified signal in order to optimize output distortions and amplifier efficiency.
PAR is the ratio of a signal’s peak power compared to its average power over a defined period of time. Complementary Cumulative Distribution Function (CCDF) is a statistical characterization that plots power level on the x-axis and probability on the y-axis of a graph. Each point on the CCDF curve shows what percentage of time a signal spends at or above a given power level. The power level is expressed in dB relative to the average signal power level.
ACPR and ACLR
Adjacent Channel Power Ratio (ACPR) and Adjacent Channel Leakage Ratio (ACLR) are terms that tend to be used interchangeably, and the differences between them are slight. ACPR is the term used to describe the power level in a channel adjacent to the transmit channel without regard for any receive filter that may be used in the communication system of interest. ACLR is a more recent term that takes into account the receiver filter used in the system of interest.
The ACLR measurement method of the latest RTSAs differs from swept techniques. Up to the limits of the maximum capture bandwidth (110 MHz for leading RTSAs), the measurement is performed on a contiguous set of time domain data containing all of the channels to be measured. Resolution bandwidths, channel bandwidths and receiver filtering are performed mathematically after the signal is digitized. The ACLR measurement in the RTSA is no different from measurements in other domains; it is merely another mathematical calculation performed on the captured signal. Since the underlying signal is the same for all measurements, this allows the RTSA to correlate measurements of modulation vs. time, CCDF and frequency domain together for enhanced troubleshooting.
Digitally Pre-Distorted Signals
Figure 3: Representative transmitter with digital pre-distortion
Whether it is a high-power satellite ground station, a multi-carrier cellular base station or even a low-power mobile system, modern transmitters employ a variety of pre-distortion techniques to reduce out-of-channel interference and optimize operating efficiency. One popular distortion reduction method is Adaptive Digital Pre-Distortion. This approach uses a sample of the transmitter’s output to calculate error vectors which are then used to pre-distort the incoming signal. To reduce analog-circuitry distortions, the signal in the chain is kept in digital format for as long as possible.
Figure 3 shows an amplifier with a low-level signal coupled from its output, down-converted and digitzed. This digitized sample is used to feed the digital signal processing circuitry, which performs analysis of the non-linearities present in the signal. These non-linear coefficients are used to alter the incoming In-phase (I) and Quadrature (Q) signals in the transmit chain. This signal, now pre-distorted and with PAR reduction applied, is fed to the amplifier after being converted back to analog by the DAC, which can be seen in the transmit chain. The resultant output signal exhibits reduced spectral distortion and lower ACLR than the signal without pre-distortion techniques. During development, prior to the availability of all parts of the design, it is common to substitute test equipment for essential elements of the design, and to substitute a workstation for the DSP system while it is under development.
Figure 4: Digital Pre-Distortion Development System
Figure 4 shows a common configuration of this type of development system. An Arbitrary Waveform Generator (AWG) is used in place of the I and Q signals and DAC, and the correction loop down-converter and ADC have been replaced with a RTSA. The I and Q vectors from the RTSA are then sent to an offline processor where pre-distortion and PAR reduction techniques are applied.
When capturing distortion products, the test instruments signal fidelity in both amplitude and phase domains are vital. The signals captured during development may contain very long sequences of specialized data, which are intended to exercise the limits of the amplifier by creating the worst-case operating scenario. These sequences may be one second in length or more, depending upon the design requirements. Top RTSAs have the ability to capture up to 1.7 seconds of I and Q data at their maximum capture bandwidth of 110 MHz. Longer captures are possible at reduced capture bandwidth.
Capturing long record lengths allows the user to examine the performance of devices in response to real-world signals. The ability to capture many packets of data is very useful in understanding design limitations, especially as they relate to changes in the signal including modulation type, number of active code channels and adaptively changing power levels .
Figure 5: Oscilloscope, logic analyzer and RTSA used in a signal path
to troubleshoot faults
The process for finding faults and troubleshooting consists of three common steps: Discover, capture and trace.
Discovering the problem can be a difficult challenge. However, leading RTSAs employ a unique and powerful spectrum-processing engine. This window into the spectral domain will capture any fault within the capture bandwidth of duration as brief as 24 microseconds with 100 percent probability of capture, ensuring that transient signals are captured on the analyzer’s screen.
Once the problem has been recognized and its characteristics understood, RTSAs allow the creation of a frequency mask trigger (FMT) to capture the signal for complete analysis in multiple domains. This is easily accomplished by referring to the spectrum display, determining where the desired signal exists and drawing a mask to trigger on any signal outside this area.
After the problem has been identified at the RF output, logic analyzers and oscilloscopes can be put to use in the baseband and IF portions of the circuit to track the problem to its source. The trigger output of the FMT can be used to trigger any other test equipment to help localize the problem.
Due to its wide capture bandwidth, deep memory and inherently correlated measurements, the RTSA is an ideal tool for the analysis and troubleshooting of wide band RF communications systems. Today’s leading RTSAs allow spectrum and vector measurements to be performed over bandwidths up to 110 MHz with high dynamic range and low residual EVM. Measurement correlation across multiple domains and frequency mask triggers greatly improve troubleshooting efforts. And new signal processing technologies enable immediate discovery of transients as brief as 24 microseconds, improving awareness of transient spectral splatter. With these capabilities, the acquisition, measurement and characterization of digitally modulated and pre-distorted RF signals are quick, efficient and accurate.
This article originally appeared in the April, 2008 issue of Portable Design. Reprinted with permission.