Microwave Engineering Europe - September 2008 - (Page 26)

26 TEST — RF AMPLIFIER PERFORMANCE Figure 4: The upper graph shows EVM versus power, while the lower graph shows the ampliﬁer gain versus input power (blue line). be easily controlled using a mouse, the analyzer’s touchpad, or remotely with a computer. In this measurement the frequency is held constant at 500 MHz while the RF input power is varied from -40 dBm to -20 dBm in 0.1 dB steps. There will be 201 amplitude steps (or points) that will take 200 ms to measure at each step. The DC bias voltage is held constant. The modulation is an 8-PSK EDGE signal, and Peak EVM is being measured with 20 measurement averages per amplitude step. Figure 4 shows the measurement results in more detail. The lower graph shows the ampliﬁer gain versus input power (blue line), which shows that the nominal gain is about 19.5 dB. It begins decreasing at about -28 to -30 dBm input power. The gain is down 1 dB at about -23.5 dBm and down 3 dB at -20 dBm. The upper graph is the EVM versus power. The red trace labeled “Distorted Plot” is the ampliﬁer EVM, and it is clear that the EVM degrades quickly as the power ampliﬁer is pushed into the gain compression region. The EVM is 1 percent or less in the linear region. It increases to about 20 percent at the 1 dB compression point and to over 40 percent at the 3 dB compression point. The upper graph shows something else as well. The green trace labeled “Baseline Plot” is the analyzer’s inherent EVM noise ﬂoor. Its EVM is about 1 percent, which is much better than the power ampliﬁer’s EVM in the compression region being measured. For this measurement, the analyzer made 4,020 accurate EVM measurements (20 measurements x 201 steps) in about 40 seconds. Figure 5 shows the Vector Signal Analyzer set-up parameters display for the EVM versus RF frequency measurement. The frequency is varied from 400 MHz to 2.5 GHz in 10 MHz steps. There will be 211 frequency measurement steps (or points) that will take about 220 ms to measure for each step. The RF input power is constant at -30 dBm. As before, the DC bias voltage is constant. The modulation, again, is an 8-PSK EDGE signal, and peak EVM is being measured with 20 measurement averages per frequency step. Figure 6 shows the measurement results in more detail. The lower graph shows the ampliﬁer gain versus frequency (blue line) showing that the gain is about 19.5 dB in the 400 to 500 MHz range and signiﬁcantly rolls off at high frequencies, about -10 dB at 2.5 GHz. The upper graph is the EVM versus frequency. The upper red trace labeled “Distorted Plot” is the ampliﬁer EVM, showing that the EVM does not degrade with frequency. The lower green trace labeled “Baseline Plot” is the analyzer EVM. This shows that inherent analyzer EVM noise ﬂoor is as good as, or much better than, the power ampliﬁer EVM. For this measurement, the analyzer made 4,220 accurate EVM measurements (20 measurements x 211 steps) in about 46 seconds. The DUT, in this case, can provide good modulation quality across its frequency range. Since an EDGE receiver is only detecting phase modulation, it can still correctly demodulate the signal even with reductions in amplitude. EDGE uses 8-PSK modulation where the constellation decision points lie around a circle of some magnitude. In this case the magnitude is decreasing with frequency but the relative phase locations of the constellation points are not changing. Thus, there is less susceptibility to EVM degradation. Although test results are not shown, the EVM of the power ampliﬁer must be characterized at a range of bias voltages to determine where EVM reaches unacceptable levels. This is particularly important for devices that will be used in mobile products. The bias level at which EVM reaches levels that prevent the receiver from correctly demodulating the transmitted signal will determine the battery voltage level at which the mobile device must shut down. Production must verify that the mobile device will still operate properly just above the deﬁned low battery threshold level. A further complication involves making EVM measurements on OFDM transmissions, Figure 5: Vector Signal Analyser set-up parameters for the EVM versus RF frequency measurement. Microwave Engineering Europe ● September 2008 ● www.mwee.com 023-024-025-026-027-028_MWEE.indd 26 2/09/08 15:52:04 http://www.mwee.com

Table of Contents Feed for the Digital Edition of Microwave Engineering Europe - September 2008

Microwave Engineering Europe - September 2008 News Contents Comment Cover Feature: Uncovering Test and Measurement’s Role in Advancing LTE and Mobile WiMAX WiMax: WiMax and LTE Need to Address Issues of Backhaul, Cost and Efficiency for Successful Deployment WiMax: Manufacturing Communications Technology Products in India Test and Measurement Mobile Platforms Evolve to Meet Future Demands Products Calendar

Microwave Engineering Europe - September 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.