This preview shows pages 1–6. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: ECE695F RFIC Prof. S. Mohammadi Nonlinearity and Distortion Let’s consider two examples Example 1 but to send this out you need an infinite bandwidth you have a QPSK transmitter LPF modulator PA BB X out X (+) (+) If you do not filter your QPSK this is what you have ° 90 ° 180 ° 180 QPSK filtered QPSK Æ you do not have infinite bandwidth this is what you get So you have to filter your signal such that it only occupies one channel if you filter your QPSK with a LPF because of limited bandwidth your phase information turns into amplitude information  146 ECE695F RFIC Prof. S. Mohammadi Now if your PA clamps the peak of your filtered QPSK you loose your information distortion due to damping changes your amplitude you loose information now you cannot correctly detect your phase in your receiver QPSK is very sensitive to amp linearity FSK or FM is not sensitive to amp linearity since you do not have instant phase variation Choice of your modulation depends on how linear your amplifier can be QPSK needs very linear amplifier FSK, FM not need linear amplifier usually class A not very efficient you can use class B, C, ,,, amplifier with much higher efficiency  147 ECE695F RFIC Prof. S. Mohammadi Example 2 your receiver receives 3 signals f ∆ + = 1 f f ∆ + = 2 2 f f 2 adjacent channels are strong interferers your signal is weak in your LNA you have nonlinearity so your strong interferers mix with each other due to LNA nonlinearity This mixing generates frequencies such as ( ) ( ) 2 1 2 2 2 f f f f f = ∆ + − ∆ + = − This is what you get at the output of LNA your signal IM3 (distortion of interferers) This distortion makes your reception difficult we have to understand distortion and nonlinearity  148 ECE695F RFIC Prof. S. Mohammadi Nonlinearity Weakly nonlinear systems (such as in receivers) Strongly nonlinear systems (such as in high efficiency transmitter) LNA class BE PA s ’ also clamping to deal with weakly nonlinear system there are two approaches power series volterra series represent transfer function using a power series power series with information about phase for strongly non linear system use envelope analysis instead of frequency domain analysis use time domain analysis * At midband frequency parasitic caps and inductors bias and bypass caps can be neglected in terms of the phase they introduce use power series * At low or high frequencies you need to know the phase of the signal that is subjected to nonlinearily use volterra series  149 ECE695F RFIC Prof. S. Mohammadi linear strong nonlinearity (cannot be modeled by power/volterra series) since at this point higher order terms become dominant in saturating the signal in S out S * most often this is what you have linear in S out S small non linearity you can write this transfer function in general by a power series L + + + = 3 3 2 2 1 in in in out S a S a S a S  150 ECE695F...
View
Full
Document
This note was uploaded on 02/19/2012 for the course ECE 695f taught by Professor Mohammadi during the Fall '09 term at Purdue University.
 Fall '09
 Mohammadi

Click to edit the document details