Downconversion and Sampling of RF Signals

1
Downconversion and Sampling of RF Signals

• Ryan Bocock Brian Otis
• EE 247 Project Presentation
• December 4, 2000

2
Objectives

• Investigate Subsampling Downconversion
• Practical limitations (i.e. noise, power)
• Explore Bandpass Downconversion Sampling as a
  possible solution
• Proof of Concept

3
Theory of Subsampling
2fbb lt fsample ltlt frf
frf
fbb
4
Effects of Filter Q on baseband SNR( Increase Q
until fBB fRF/2Q )
5
Switch Thermal NoiseSwitch BW 1GHz,
fclock10MHz, Csample100fF
Integrated Thermal Noise Calculated40nV2 Simulat
ed (Closed) 35.11nV2 Simulated (Clocked) 34.16nV2
6
Power Constrained Switch Noise Optimization
f
Cclock
Csample
Rds,on
7
Example fc2GHz, fbb5MHz
8
Bandpass Downconversion Sampling

• Requirements
• Downconversion
• Anti-alias filtering
• Sampling
• Goals
• High level of integration
• Low power consumption

9
Weighted Integration
1
fclock
-1
2fbb
A
D
frf

• Implements BP downconversion filter
• fclockfcenterfrf
• Integrate over N clock periods
• N lt fclock / 2fbb

10
Single Weighted Integration
Convolution of pulse with signal produces sinc
filter frequency response
Sample
time
freq
Integrate
Reset
11
Simplified Bandpass Downconversion Sampling
Schematic
12
Matlab Baseband Spectrum Comparison
13
Spectre Frequency Response, N3
14
Effects of Increasing Integration Samples
N3
N30
N300
15
Parallel Weighted Integration
16
Matlab Baseband Spectrum Comparison
17
Subsampling Results
18
BP Downconversion Sampling Results