Design of RF CMOS Low Noise Amplifiers Using a Current Based MOSFET Model Virgínia Helena Varotto Baroncini Oscar da Costa Gouveia Filho
OUTLINE 1. Introduction 2. MOSFET Model 3. High-Frequency Noise Model 4. LNA Analysis 5. LNA Design Example 6. Conclusion
Introduction Submicrometer CMOS technology allows the integration of RF circuits. Low voltage and low power operation → moderate inversion Model valid from weak to strong inversion
MOSFET MODEL S D B G I(V G,V S ) I(V G,V D ) I F = forward current I R = reverse current
Normalized currents where are the normalized currents andis the normalization current
Operation Regions of the MOS transistor irir ifif forward saturation i f > 100 i r reverse saturation i r > 100 i f strong weak moderate strong moderate weak triode
Small signal parameters Transconductances Capacitances
High- Frequency Noise Model S vRg S ig S id RgRg C gb C gs g ms V sb g m V gb B B S S G D
Channel Thermal Noise ifif S id (A 2 /Hz)
Induced Gate Noise
LNA Analysis LSLS LgLg LdLd V in VbVb V DD M1M1 M2M2 Cascode LNA with inductive source degeneration
Impedance Matching LgLg LsLs C gb C gs g mb V sb g ms V gs Z in Z1Z1 Z 1 can be viewed as the parallel of a resistor R with the capacitance Cgs
Simplified small signal model for the LNA matching is achieved simply by making the real part of Z in equal to the source resistance and its imaginary part equal to zero.
Noise Figure LNA small-signal model for noise calculations Definition The noise figure can be expressed as a function of i f
Noise figure versus W/L for several inversion levels at 2.5 GHz
LNA Design Example Resonance frequency2,5 GHz Supply voltage2,5 V Length 0,35 m Source resistance 50 Noise Figure< 2dB LNA Design Parameters
1. Choice of the inversion level i f =35 Procedure
2. L s for impedance matching
3. Transistor width for minimum noise figure Noise figure versus W/L for several inversion levels at 2.5 GHz
4. L g to satisfy the resonance frequency 5. L d to adjust the gain and the output resonance frequency
LNA Design Results W/LWIDID RsRs LsLs LgLg LdLd m 4,1 mA 50 0,7 nH7,6 nH2,5 nH LsLs LgLg LdLd RSRS CLCL R Ld M1M1 M2M2 V out V in + V bias V DD
Simulation results Input impedance
Noise Figure
Conclusions The main advantage of this methodology is that is valid in all regions of the operation of the MOS transistors; It is possible to move the operation point of RF devices from strong inversion to moderate inversion taking advantage of higher gm/ID ratio, without degrading the noise figure;