ET in PA

1/3/12
www.eetimes.com/General/PrintView/4233749
Understand and characterize envelope-tracking power amplifiers

Gerard Wimpenny, Nujira Ltd. & Member of OpenET Alliance 1/3/2012 8:50 AM EST
The process of designing traditional fixed supply power amplifiers (PAs) has been well established for many years. Well defined metrics for performance assessment exist, and the PA designers job is simply to design a PA with the best set of performance metrics. In reality of course this is far from a simple task, but at least the assessment criteria are well established and well understood. For envelope-tracking (ET) PAs, the situation is more complex and requires the use of more sophisticated characterization techniques. ET Basics The objective of envelope tracking is to improve the efficiency of PAs carrying high peak to average power (PAPR) signals. The drive to achieve high data throughput within limited spectrum resources requires the use of linear modulation with high PAPR. Unfortunately, the efficiency of traditional fixed supply PAs when operated under these conditions is very poor. The efficiency of an ET PA is improved by varying the PA supply voltage in synchronism with the envelope of the RF signal. The PAs fundamental output characteristics (power, efficiency, gain, phase...) now depends on two control inputs (RF input power and supply voltage) and may be represented as 3D surfaces. In a typical envelope tracking system, the supply voltage is dynamically adjusted to track the RF envelope at high instantaneous power. Here, the PA operates with high efficiency in compression. Its output characteristics are primarily determined by the instantaneous supply voltage. Conversely, when the instantaneous RF power is low, the supply voltage is held substantially constant and the PA output characteristics are primarily determined by the instantaneous input power (linear region). A transition region in which both supply voltage and input power influence the output characteristics exists between these two extremes (see Figure 1)
1/9
1/3/12
Figure 1 Instantaneous efficiency vs supply voltage Isogain shaping ET PA Linearity A simple quasi static (i.e. memoryless) behavioral model of a PA may be constructed if its AM/AM and AM/PM characteristics are known. These characteristics, along with other key PA metrics such as power and efficiency are profoundly influenced by the mapping between instantaneous RF envelope and applied supply voltage. In an ET system this mapping is determined by the contents of a shaping table in the envelope path (see Figure 2)
Figure 2 Envelope Tracking PA System A mapping of particular interest is ISOgain shaping, in which the instantaneous supply voltage is chosen to achieve a particular constant PA gain (see Figure 3).
Figure 3: RF O/P power - Supply voltage mapping - ISOGain shaping With this mapping, the ET PA system achieves low AM-AM distortion despite operating in compression over much of the envelope cycle as shown in Figure 4. The equivalent trajectory for fixed supply operation is also shown in Figure 4 from this it is apparent that ET can actually be used to linearize a PA, reducing ACPR and EVM.
2/9
1/3/12
Figure 4: ET PA gain characteristics - Isogain shaping The system trade-off associated with using the shaping table to linearize the PA is a small loss of efficiency (compare Figure 1 and Figure 5) for a substantial improvement in linearity (compare Figure 4 and Figure 6). The choice of shaping function also has a strong influence on the bandwidth requirement of the envelope path. A smooth transition between the linear and compressed regions results in a lower bandwidth requirement for the envelope amplifier for modest (1-2%) loss in system efficiency.
Figure 5: ET PA Efficiency - Optimum efficiency shaping
3/9
1/3/12
Figure 6: ET PA gain characteristics - Optimum efficiency shaping When designing a fixed supply linear PA, a great deal of attention must be paid to the achieving adequate linearity characteristics at maximum output power. Many factors influence the linearity (e.g. fundamental technology characteristics, biasing and RF matching) and it is up to the PA designer to achieve the best trade-off between efficiency and linearity. For an ET PA, things are different, as the linearity of an ET PA in the compressed region is no longer a self contained PA parameter. The PA still has to be linear in the low power, low voltage region, but at higher powers the AM linearity constraint is removed and the PA can be designed for optimum ET efficiency without regard to AM linearity. Unlike AM distortion, phase distortion is not directly controlled by the envelope shaping table. However, it is observed that many PAs actually show reduced PM distortion when operated in ET mode. As a result of this self linearization, it is possible to push harder into compression at signal peaks than with a fixed supply amplifier, allowing increased output power for given linearity. Figure 7 shows measured ACLR and EVM performance for a PA operated in Fixed Supply and ET modes. In this example the PA output power for -40dBc ACLR is 2dB higher in ET than fixed supply mode.
4/9
1/3/12
Figure 7: ET PA / Fixed supply PA linearity comparison
Characterization Techniques The standalone performance of ET PAs cannot be measured unless the shaping table is first defined. This requires measurement of the PAs fundamental characteristics (Pout, Efficiency, Gain, Phase) over the full range of supply voltage and input power. In principle this characterization could be carried out using a CW network analyzer and a variable DC supply, but results are typically poor due to thermal effects, ranging errors and drift in phase measurements. It is also far too slow to allow load pull techniques to be used. An alternative approach is to use a pulse characterization using ATE controlled standard test equipment. This avoids the need for a high bandwidth low impedance supply and is sufficiently fast for Load Pull to be viable, but has the drawback that it is difficult to make accurate phase measurements. The last approach is to use real waveforms and to vary the shaping table to allow all combinations of input power and supply voltage to be measured. This requires a supply modulator, but is very fast, allows accurate phase information to be gathered and can also be used to characterize memory effects.
5/9
1/3/12
Figure 8: ET PA Characterization bench (Click figure to download larger version in PDF) A basic ET PA characterization can be used to create a quasi-static (i.e. memoryless) data model of the PA having output power, phase and efficiency as outputs and input power and supply voltage as inputs. Once the shaping table is defined, the model can be used to predict PA system performance parameters such as ACPR, EVM and efficiency for standard test waveforms. In addition to being used for PA device level characterization, the same hardware can be used for direct verification of PA system performance using the defined shaping table (see Figure 9).
Figure 9: ET PA AM/AM and AM/PM after shaping table definition For higher bandwidth waveforms, PA memory effects can be a significant source of non linearity. The PA output parameters (AM, PM, efficiency) now depend on time (i.e. signal history) in addition to instantaneous input power and supply voltage. Memory effects show up in the PA characterization as a broadening of the AM/AM and AM/PM characteristics and can result from electrical time constants in input or output bias circuits, thermal time constants associated with local die heating, or technology specific charge storage effects. Efficiency Optimization The statistics of typical high Peak to Average Power Ratio (PAPR) signals are such that an ET PA typically spends most of its time operating with relatively low supply voltage, with only occasional high voltage excursions on high power peaks. It makes sense therefore to optimize the PA matching to achieve best efficiency with the target high PAPR signals rather than simply designing for best efficiency at peak power / max supply voltage, as would be the case for a fixed supply PA. It can be seen from Figure 10 that the PA matching should be altered to increase efficiency around the peak of the signal probability density function, even if this necessitates a slight compromise in the peak power efficiency.
Figure 10: Influence of signal statistics on Efficiency of an envelope tracking PA To fully optimize the efficiency of an envelope tracking PA the device characterization can be extended to include sweeping the load impedance (fundamental or harmonic load pull) in addition to input power and supply voltage. Analysis of the large dataset produced by such a characterization can be automated (e.g.
6/9
1/3/12
using MATLAB) to predict the average PA efficiency when operating with a specific set of ET parameters. For example, using this characterization methodology it is possible to predict how a PAs average efficiency varies with shaping function, output voltage swing range, back off from maximum power and waveform statistics when operated in ET mode (see Figure 11).
Figure 11: ET average Efficiency and PA Output power load pull contours (Click figure to download larger version in PDF) ET PA parameter variation sensitivity It is commonly expected that the performance of ET PAs over temperature will be poorer than their fixed supply counterparts. In fact the reverse is true - unlike a fixed supply PA, an ET PAs performance is much more sensitive to changes in the supply voltage characteristics than to changes in gain of the RF chain driving the PA. As the characteristics of the supply voltage can be much better controlled over temperature than RF gain, little variation in PA linearity is observed for extreme temperature variations (See Figure 12).
Figure 12: ET PA ACLR vs Temperature (105C range) In a handset environment, the load impedance presented to the PA is not well controlled due to reflections
7/9
1/3/12
from nearby objects, which can result in the PA having to work into load mismatches as high as 3:1 VSWR. The ET PAs self linearization principle previously described is also applicable under high VSWR conditions and this can result in significantly improved EVM and ACPR performance compared with the same PA operated in fixed supply mode (see Figure 13)
Figure 13: Comparison of EVM vs backoff for ET and fixed supply PAs operating into 3:1 VSWR load mismatch Summary The system efficiency benefit of operating a PA in ET mode is widely known. However, it also offers other useful system benefits, such as increased output power, improved operation into mismatched loads, and insensitivity to temperature variations. In contrast to fixed supply PAs, the performance of an ET PA is not self contained and requires substantially more data to be gathered to predict system performance. This requires use of a test environment which allows the supply voltage in addition to input power to be swept. A key aspect is the definition of the shaping table which defines the relationship between supply voltage and RF power. This defines many key PA metrics. Once the shaping function is defined, efficiency and linearity can be directly measured using an appropriate system characterization bench. About the author Gerard Wimpenny Chief Technology Officer, Nujira Ltd. & Member of OpenET Alliance Gerard has over 20 years of RF and Signal Processing experience, and has been responsible for strategic R&D, design process definition, and top level technical support for business development activities. He has been instrumental in delivering numerous wireless products to semiconductor vendors, infrastructure and handset manufacturers and holds an MA degree from Cambridge University. If you liked this article... Head to the RF and Microwave Designline homepage for the latest updates in RF and microwave. Sign up for the weekly RF and Microwave Designline Newsletter to be delivered to your mailbox
8/9
1/3/12
with the latest highlights from the site.
9/9

ET in PA

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ET in PA

Uploaded by

Copyright:

Available Formats

1/3/12

Understand and characterize envelope-tracking power amplifiers

Figure 5: ET PA Efficiency - Optimum efficiency shaping

Figure 7: ET PA / Fixed supply PA linearity comparison

with the latest highlights from the site.

You might also like