EE314 Stanford Lectures On RF

EE314:
CMOS RF Integrated Circuit Design
Introduction to Wireless Communication systems

Stanford University
Hamid Rategh
Hamid Rategh Stanford University EE314 HO#1 1

Course Staff
Instructor: Dr. Hamid Rategh
Email: [email protected]
Office Hours: MW 2:15-3:15PM @ CIS-126;
Phone: 725-8313
TA: Mehdi Jahanbakht and Deji Akinwande

Office Hours @ Packard 106

Sunday 3:00 – 3:30 pm (for SCPD students only)
Sunday 3:30 – 5:00 pm
Thursday 5:00 – 6:00 pm
Course Administrator: June Wang
Office: CIS-203
Phone: 725-3706

TA announcements
Class URL is: http://eeclass.stanford.edu/ee314/
All students should register on class website to access handouts
and to stay in touch with any announcements from the instructor
or TA's
The bulletin board on the class website will be supported by the
TAs for exchange of information
For the SCPD OHs, (per encouragement from SCPD) we will be
experimenting with instant message chatting (instead of phone
calls) in an attempt to provide better TA OH support to more
SCPD students.
We will send instructions to the SCPD students in a few days.

Course snapshot
Primary text
“The Design of CMOS RF Integrated Circuits”, T.
Lee, Cambridge, 2004 (Second Edition)
Recommended text:
“RF Microelectronics”, B. Razavi, Pearson
Education, 1997
Grading
Homework 30%, Project 30%, Final 40%
Prerequisite: EE214
Course Topics
Introduction to wireless communication systems
Receiver architectures
Review of passive networks
Available passives in IC
RLC networks and tune circuits
Impedance transformation techniques
Transmission lines
Review of Distortion and circuit non-linearity
IP3
1dB compression point
AM to PM distortion
Noise
Noise sources in passive and active circuits
Classical noise theory
LNA design

Course Topics (Cont.)
Mixers
frequency conversion techniques
Passive and active mixers
Oscillators
Topologies (Ring, Colpitts, VCO, Quadrature, …)
Phase noise
Frequency synthesizers and Phase-locked loops (PLL)

Integer-N
Fractional-N
Power Amplifiers
Different classes of operation (A, B, C, D, …)
Linearization techniques

Block diagram of a wireless transceiver
Front-end
De-modulator
Receiver
Base-band
Duplexer/ LNA, Mixer, VCO, PLL,…
Signal
Switch
Processing
Front-end
Modulator
Transmitter
PA, mixer,…
In the transmit path, the base-band processor sends out the coded and
compressed digital bits, which are then modulated and up-converted to
the transmit frequency and finally amplified by the front-end module and
transmitted via antenna
In the receive path the received signal from the antenna is amplified
and down converted to either base-band or some other intermediate
frequencies before it is processed with the base-band signal processor
In this course we will be only looking at the front-end receiver and
transmitter

Half-duplex systems
Front-end
Switch Receiver
Front-end
Transmitter
In a half-duplex system the transceiver either

transmits or receives at any given time
Such as: Walki-talki, GSM
The antenna is switched between transmitter and
receiver
Typically antenna switches have 1-2dB insertion
loss and provide about 40dB of isolation between
the two ports

Full-duplex systems
Front-end
Duplexer Receiver
Front-end
Transmitter
In a full duplex system the transceiver can transmit and receive

simultaneously
Such as: WLAN, CDMA, WCDMA
How can we transmit and receive at the same time?
Transmit at one frequency and receive at a different frequency
Duplexers are used to share the same antenna for transmit and
receive.
The insertion loss of isolators is generally in the order of 1-2dB
About 30-40dB isolation between the two ports

Multiple access
From basic communication course:

Different users (transmitters or receivers) share
the same medium (i.e., air) by:
Transmitting at different frequencies (i.e., frequency
division multiple access (FDMA))
Transmitting at different time slot (i.e., Time division
multiple access (TDMA))
Transmitting with a different code (i.e. code division
multiple access (CDMA))

Multiple channels and channel selection
Undesired 60dB filter
channel
B f
In a typical wireless system, the frequency band is divided into
smaller divisions called “channel”
Consider a system with f=1.9GHz and B=1MHz
Assume we intend to use a filter at the carrier frequency to select
the desired channel from the adjacent channel which is 60dB
stronger
If we had to use a second order LC filter: the Q should be in
excess of 107 to attenuate the adjacent channel 10dB below the
desired channel
Such a high Q is very difficult (if not impossible) to achieve even
with crystal/SAW filters at GHz frequencies.
How do we practically select the desired channel?

Heterodyne Receiver
A typical heterodyne receiver

comprises of an optional front-end
amplifier two filters and a mixer for
frequency conversion.
ωmix = ω RF ± ω LO
If we select the IF to be the
difference of the LO and RF
frequencies then the receiver is
called superheterodyn
ω IF = ω RF − ω LO
The main reason for frequency conversion is to make channel selection feasible
with practical filters
The required Q of the filter is inversely proportional to the center frequency
If fIF=10MHz and fRF=2GHz, then the required Q for the channel select filter will 2 orders of
magnitude less when channel selection done at IF, instead of RF

Image frequency in heterodyne receivers
ωLO
Consider a superheterodyne receiver.
ω IF = ω RF − ω LO ⇒ ω RF = ω LO ± ω IF
Therefore there are two input frequencies which down-convert to the same IF
frequency. One is the desired channel and the other is called “image frequency”
ω RF − ωimage = 2ω IF
Image reject filters are used to suppress the image frequency

From channel selection discussion we learned that the lower the IF the lower the
required Q of the filter. But is this also good for image rejection?

Selection of IF frequency
A large IF frequency makes image rejection simple

and channel selection difficult
A Low IF frequency makes image rejection simple
and channels selection difficult
Is there a way to break this relationship?

Heterodyne receiver with dual IF
IF1 IF2
IF2 < IF1

Channel selection is done in two stages
Therefore the requirements for each of the channel
select and image reject filters is relaxed
But now we need more filters.
Filters are generally off-chip components and are not
desirable for fully integrated solutions

Homodyne (zero IF) receiver
ω RF = ω LO ⇒ ω IF = 0 ⇒ ωimage = ω RF
In a homodyne receiver there is no image

frequency
There is no need for image-reject filter
Channel select filter is simply a low pass filter
Murphy says there is no free lunch! What is
the catch?
Zero-IF receiver challenges
z Susceptibility to flicker noise

Weaver image-reject receiver
vout = vin [cos(ω1t ) cos(ω2t ) − sin(ω1t ) sin(ω2t )] = vin cos[(ω1 + ω2 )t ]

If ω1+ ω2= ωRF then weaver receiver operates like a zero IF receiver and no
image reject filter is needed
Conversion to DC is done at the second stage where ω2<< ωRF and therefore LO
and interferer leakage is less of an issue
Unlike homodyne receiver the received signal is already amplified well above the
flicker noise level before converted to DC and therefore flicker noise is also less
of an issue in weaver receiver
The two filters are simple low pass filters to suppress ωRF+ ω1.Why do we need
to that?

Reading material
Razavi: chapter 5

EE314 Stanford Lectures On RF

Uploaded by

Copyright:

Available Formats

EE314 Stanford Lectures On RF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EE314 Stanford Lectures On RF

Uploaded by

Copyright:

Available Formats

EE314:

CMOS RF Integrated Circuit Design

Introduction to Wireless Communication systems

Hamid Rategh Stanford University EE314 HO#1 1

 Office Hours: MW 2:15-3:15PM @ CIS-126;

 TA: Mehdi Jahanbakht and Deji Akinwande

 Office Hours @ Packard 106

Hamid Rategh Stanford University EE314 HO#1 2

Hamid Rategh Stanford University EE314 HO#1 3

Hamid Rategh Stanford University EE314 HO#1 5

 Passive and active mixers

 Frequency synthesizers and Phase-locked loops (PLL)

Hamid Rategh Stanford University EE314 HO#1 6

Hamid Rategh Stanford University EE314 HO#1 7

 In a half-duplex system the transceiver either

Hamid Rategh Stanford University EE314 HO#1 8

 In a full duplex system the transceiver can transmit and receive

Hamid Rategh Stanford University EE314 HO#1 9

 From basic communication course:

Hamid Rategh Stanford University EE314 HO#1 10

Hamid Rategh Stanford University EE314 HO#1 11

 A typical heterodyne receiver

Hamid Rategh Stanford University EE314 HO#1 12

 Consider a superheterodyne receiver.

 Image reject filters are used to suppress the image frequency

Hamid Rategh Stanford University EE314 HO#1 13

 A large IF frequency makes image rejection simple

Hamid Rategh Stanford University EE314 HO#1 14

 IF2 < IF1

Hamid Rategh Stanford University EE314 HO#1 15

 In a homodyne receiver there is no image

z Susceptibility to flicker noise

vout = vin [cos(ω1t ) cos(ω2t ) − sin(ω1t ) sin(ω2t )] = vin cos[(ω1 + ω2 )t ]

Hamid Rategh Stanford University EE314 HO#1 18

Hamid Rategh Stanford University EE314 HO#1 19

You might also like

Office Hours: MW 2:15-3:15PM @ CIS-126;

TA: Mehdi Jahanbakht and Deji Akinwande

Office Hours @ Packard 106

Passive and active mixers

Frequency synthesizers and Phase-locked loops (PLL)

In a half-duplex system the transceiver either

In a full duplex system the transceiver can transmit and receive

From basic communication course:

A typical heterodyne receiver

Consider a superheterodyne receiver.

Image reject filters are used to suppress the image frequency

A large IF frequency makes image rejection simple

IF2 < IF1

In a homodyne receiver there is no image