elen 701 rf & microwave systems · pdf fileelen 701 rf & microwave systems engineering...
TRANSCRIPT
ELEN 701 RF & Microwave Systems Engineering
Lecture 6October 25, 2006
Dr. Michael ThorburnSanta Clara University
Announcements
• Today– Collect HW#2– Collect late HW#1
• HW#1 to be returned next week (w/ apologies for delay) – so late submissions will not be accepted after I have returned assignment to rest of class
• Last 15 minutes of class today – visit from department administration
Transmitter System Analysis and Design
• Part I (Lecture 6 - October 25) • Introduction• Transmission Power and Spectrum• Modulation Accuracy
– Error Vector Magnitude and Waveform Quality Factor
– Influence of Intersymbol or InterchipInterference to EVM
– Influence of Close-in Phase Noise of Synthesized LO to EVM
– Carrier Leakage Degrading the Modulation Accuracy
– Modulation Accuracy Degradations Resulting from Other Factors
• Degradation Due to I and Q Imbalance• Nonlinearity Influence on EVM• Impact of In-Channel Bandwidth Noise• Modulation Error Resulting from Reverse
Modulation of LO– Total EVM and Waveform Quality Factor
• Part II (Lecture 7 - November 1)• Adjacent and Alternate Channel Power
– Low-Pass Equivalent Behavioral Model Approach
– Multitone Techniques– ACPR of Cascaded Stages in Transmitter
Chain• Noise-Emission Calculation
– Formulas for Noise-Emission Calculation– Some Important Notes in Noise-Emission
Calculation• Output Noise of an Attenuator• Output Noise Floor of Device or Transmitter• Product of Noise Factor F and Gain g Greater
than One• Input Noise Floor of a Device
– Noise Expressed in Voltage• Some Important Considerations in System
Design– Comparison of Architectures– Transmitter Chain Gain Distribution and
Performance– AGC and Power Management
Superheterodyne Full-Duplex Architecture Configuration
Transmitter Section
Introduction
• Transmitter is the companion of the receiver in wireless mobile stations– They operate simultaneously in Full-Duplex
systems– May run in different time slots in Half-Duplex
systems– Architecture of a transmitter can also be:
• Superheterodyne• Direct Conversion• Band-pass Sampling
Features of Transmitter• Transmitter architecture does not employ an IF SAW
filter• The signal processed and amplified in a transmitter is
deterministic since it is generated in the local digital baseband.
• The signal level is much higher than in a receiver• Important Parameters Include:
– Maximum Output Power– Modulation Accuracy, EVM or Waveform Quality Factor– Adjacent Channel Power– In-band and Out-of-band noise/spurious emissions– Nonlinearity of Power Amplifier (and driver amplifier)– Automatic Level Control
Superheterodyne Full-Duplex Architecture Configuration
Transmitter Section Only SAW Filter is post Driver Amp
Receives Signals over Full Bandwidth
Generates SignalFrom BBMultiple
Input Filters
Transmitter Chains
• Example from Page 353 provides ACPR, Noise in Rx Band, etc.
Function Block DAC Atten LPF Mod RF VGAImped Match RF VGA Driver SAW PA Coup Duplex Diplex ANT
Power Gain -8.00 0.00 5.00 4.17 -1.00 12.00 8.00 -4.00 27.00 0.40 -2.50 -0.50Pout (dBm) -15.57 -23.57 -23.57 -18.57 -14.40 -15.40 -3.40 4.60 0.60 27.60 28.00 25.50 25.00 25.00
Cascaded OIP3ACPR (dBc)
Next week we will add Cascaded OIP3 Expression
Parameters of Interest
• Today!– Parameters to Characterize Transmitter
• EIRP (or ERP)– Combination of Transceiver and Antenna!
• Error Vector Magnitude and Waveform Quality Factor
EIRP
ComponentLosses
ComponentLosses
PAOutput Power
RepeaterOutput Losses
AntennaDirectivity
AntennaLosses
AntennaPointing
EIRP EIRP
Antenna GainRepeater Output Power
OMJTC
AntennaRepeaterPA Isolator
Duplxr
Key Component: Driver Amp / Power Amp
Driver /PADriver /PA
Freq
Pow
er
EPC
Freq
Pow
er
• Driver amplifier provides medium power amplification and gain control
• Linearizer is sometimes needed to improve transponder linearity when high power amplifier (HPA) is operated in the non-linear region
• HPA provides the final stage of high power amplification
– Solid State Power Amp (SSPA)
• HPA typically is the main contributor to signal distortion due to its non-linear characteristics when operated near saturation
PA Transfer Characteristics
LinearRegion
Non-linearRegion
Phase Shift vs. Input Power Back Off
-5
5
15
25
35
45
55
-40 -30 -20 -10 0 10
IBO (dB)
Phas
e Sh
ift, D
egre
es
LinearRegion
Non-linearRegion
PA operatedin linear region
Freq
Pow
er
Freq
Pow
er
C/3IM = Carrier to 3rd order IM ratio
PA operatednear Saturation
Freq
Pow
er
Freq
Pow
er
Transmission Power and Spectrum
• Transmitter Power– Key component of EIRP– Directly dependent of Power Amplifier
Capability– Power measured in the frequency domain in
terms of the integrated power over a bandwidth
• Think of a Raised Cosine Filter• See Figure 5.1 on page 313.
Next Parameter
• Error Vector Magnitude – Measure of Modulation Accuracy
Modulation Accuracy
• The modulation accuracy is represented by error vector of modulation– EVM is the difference between the actual symbol location
and the theoretical symbol location on the modulation vector constellation diagram.
a_i
a_q(-1,1) (1,1)
(-1,-1) (1,-1)
Phase Shiftpi/4, 3pi/4, -3pi/4, -pi/4
OQPSK modulationconstellation
)()()( 111 kekaka +=′
Residual Error Vector
Error Vector Magnitude and Waveform Quality Factor
• Error Vector Magnitude (EVM)– Defined as the mean
square error between the samples of the actual and the iidealsignals, normalized by the average power of the ideal signal
{ }{ }
{ }{ }
21
21
21
21
21
211
)(
)(
)(
)()(
⎥⎥⎦
⎤
⎢⎢⎣
⎡=
⎥⎥⎦
⎤
⎢⎢⎣
⎡ −′=
kaE
keEEVM
kaE
kakaEEVM
Waveform Quality Factor
• Waveform Quality Factor– Defined as a
correlation coefficient between the actual waveform Z(t) and the ideal waveform R(t)
2
1 1
22
2
1
11
EVM
ZR
ZR
M
k
M
kkk
M
kkk
+≈
=
∑ ∑
∑
= =
=
∗
ρ
ρ
Note: Appendix 5A provides the proof for relationship between Waveform Quality Factor and Error Vector Magnitude
Waveform Quality Factor
• What do we need to remember?– The modulation accuracy is represented by error
vector of modulation• EVM is the difference between the actual symbol location
and the theoretical symbol location on the modulation vector constellation diagram.
– Waveform Quality Factor• Defined as a correlation coefficient between the actual
waveform Z(t) and the ideal waveform R(t)– Perfect Modulation Accuracy
• EVM=0• WQF=1
Influence of Intersymbol or Interchip Interference to EVM
• The modulation accuracy of a modulated RF/IF signal can be degraded when the signal passes through a nonideal filter– Rationale
• A modulation signal consists of symbols• The symbol waveforms are distorted due to the filter group
delay distortion and magnitude response ripple• One symbol generates interference in the adjacent (and
possibly other) symbols– ISI = Intersymbol Interference– ICI = Interchip Interference
∑∞
−∞=
Δ=k
ICIISI kIEVM )(2
Influence of Close-in Phase Noise of Synthesized LO to EVM
• Another main contribution to the degradation of the modulation accuracy is– The close-in phase noise of the LO synthesizers
• in the modulator • and in the upconverter
( )
( )tte
etatetata
n
tj n
22)(
)()()()(
φ
φ
=
=+=′
The statistical average of this is the autocorrelation function P of the phase noise.
Influence of Close-in Phase Noise of Synthesized LO to EVM
• N = average phase noise• P = autocorrelation function
looprsynthesize
N
Nphase BWPphase
_10102 ⋅⋅≅
∑=
=N
kkNphaseNphase PEVM
1, N=number of synthesizers
Additional Factors
• Carrier Leakage Degrading the Modulation Accuracy– The DC offset in the
BB I and Q channels will cause carrier leakage and it will degrade the modulation accuracy
• Others– Degradation due to I
and Q imbalance– Nonlinearities– Impact of In-Channel
Bandwidth Noise– Modulation Error
Resulting from Reverse Modulation of LO
• Reflected harmonic of transmission signal
Total EVM and Waveform Quality Factor
∑=k
kTOTAL EVMEVM 2
Assumption is that all of these modulation distortion terms are uncorrelated!
Homework (due Nov 8)
• Develop a Spreadsheet (or Program) to calculate the Signal Level, Cascaded NF, Cascaded OIP3 (C/3IM) for the cascaded set of components comprising your transmitter.– Give yourself enough flexibility so that you can add components
(passive or active) to your line up– Use example in text on page 353 to validate your equations
• The example on page 353 gives the noise power in the receive band to be ~ -173.6 dBm/Hz. Assume the receiver is modeled by the example on page 297. What would the degradation in Noise Figure be due to the transmit power in the receive band?