Transcript Radar related progress in RF technology over the past decade

Radar related progress in
RF technology over the
past decade
(A short trip down Memory Lane)
PW van der Walt
Reutech Radar Systems
University of Stellenbosch
Introduction


We’re on the threshold of the last year of the decade – good time to look back!
The first decade of the 21st century has witnessed the maturation of important
high speed device technologies





During the past decade devices using these technologies have matured, driven
mainly by cell phone and other communications systems



Hetero-junction Bipolar Transistors and HBT ICs
HEMPT and pHEMPT devices and ICs
LDMOS power devices
Ever smaller and faster silicon CMOS devices
In many cases radar applications benefit because the devices also meet radar
requirements
Looking back, it has been an exciting 10 years!
In this paper a brief outline is given of some of the more spectacular advances in
RF devices that have made them useful in the implementation of radar systems
Contents

Frequency synthesis



DDS
PFD
Signal conditioning


Amplification
Frequency translation




Modulators, demodulators and mixers
Data conversion
RF Power
Packaging microwave components


PCB materials
PCB design
Frequency synthesis

DDS




A DDS consists of a phase accumulator (= a wrap around counter), a
sine look-up table and a digital to analog converter (DAC)
The direct digital synthesizer has become a key element in frequency
synthesizers
Radar requirements are strict in terms of bandwidth, phase noise and
spurious signals. The bandwidth and spurious responses of earlier DDSs
limited the application of early generation DDSs in radar systems.
DDS spurious caused by

DAC harmonics that alias back to baseband


Mitigated by improved linearity
Phase truncation spurs when the frequency tuning word msb’s > DAC
resolution

Situation improves with ADC resolution
DDS Application

Because the DDS output is a sampled signal,
aliasing of harmonics occurs
DDS wide band spurious can be quite high
 DDS narrow band spurious can be quite low
 A PLL acts as a tracking narrow band filter and is
the ideal companion for a DDS


The following slide shows the preferred
application of a DDS in a wide-band frequency
synthesizer for radar
DDS Application
DDS then and now

2000:

State of the art represented by AD9854

300 MHz clock, 12 bit ADC, 48 bit FTW




Falls short of many radar requirements
Digital Phase Frequency Detectors (PFD) fall short on phase noise
2008:

Single IC state of the art represented by AD9912

1000 MHz clock, 14 bit ADC, 48 bit FTW



80 dB narrow band SFDR at 100 MHz
86 dB narrow band SFDR at 400 MHz
Backed by a matching PFD
The DDS qualifies excellently for radar applications



Agile LO Synthesizers
Chirped pulse generators
FMCW sources with excellent chirp linearity
Frequency Synthesis

PFD


The phase-frequency detector is at the heart of a PLL
The first decade of the 21st century saw the introduction of a low noise
phase-frequency detector by Hittite, bettering its predecessors by about
20 dB in one spectacular leap





The HMC439 has a noise floor of -140 dBc/Hz at 1.3 GHz
This is comparable to the noise performance of well-designed analogue
synthesizers for radar
The PFD made possible the use of simple PLL architectures for high
performance radar frequency synthesizers
And it happened in the past decade
As a case in point, RRS now builds a synthesizer optimised for
its X-band pulsed MTD radars on two PC boards with phase
noise performance better than that of analogue synthesizers
costing R1 million a decade ago
Amplification

Progress with amplifiers was steady rather than
spectacular




Gain block performance has improved with lower noise
figures and higher third order intercept points (IP3)
Gain stability has reached an impressive ±0.2 dB over -40º to
85º C temperature range
Bandwidths of 20 GHz
Modern HBT based amplifiers have extremely low phase
noise at high drive levels, as exemplified by the measured
phase noise at 1 GHz of a ×10=×2×5 frequency multiplier
implemented with a new generation of amplifiers.
1 GHz Multiplier phase noise
Frequency Translation

Modulators can produce
single sideband RF output
signals up to about 4 GHz
from baseband IQ inputs
from a single integrated
circuit which incorporates
the wideband LO
quadrature polyphase phase
shift filter
Modulators





At the turn of the century, baseband bandwidths were limited to
250 MHz and noise floors were at a level of about -152 dBc/Hz
Today modulators with baseband bandwidths of 700 MHz and
noise floors of -158 dBc/Hz
Carrier suppression without trimming can exceed -50 dBc and
sideband rejection is better than 40 dB
The translinear multiplier cells far outperform diode mixers
regarding spurious performance
These devices have proved to be high performance and versatile
elements in radar frequency synthesizers and transmitter exciters

With demodulators to match
Frequency translation

Another welcome component is the balanced
diode mixer whith incorporated LO amplifier,
reducing external LO drive requirements to less
than 0 dBm, while internally driving the mixer at
a level of 20 dBm
This comes as a blessing to everyone who has had
the pleasure of suppressing leakage from high level
LO signals
 At the same time offering mixer/demodulator with
very large dynamic range

Data Conversion


The performance of ADCs has improved by
leaps and bounds during the past 10 years
2000:
80 MS/s, 10 bits, 300 MHz bandwidth (AD)
 50 Ms/s, 12 bits, 300 MHz bandwidth


2009:


1 GS/s, 12 bits, 2.1 GHz bandwidth (TI)
Rate of increase in speed:

1 decade/decade!
Bad and Good News

The Bad News for RF engineers
The digital domain is encroaching on their analogue
world
 One of these days we will be left with nothing but
antennas and LNAs!


The Good News for RF engineers
Mixed circuit technology has become an RF
problem
 PCB design is now done by a multi-disciplinary team

Power Transistors

Prior to about 2000, bipolar transistors were the only
devices on offer for L-band radar transmitters



LDMOS transistors are now the devices of choice




Up to 200 W peak power per device
7 dB power gain
Up 500 W of peak power per device (IFF)
12 dB power gain
50 V supply
These devices dramatically impact on the complexity of
radar transmitters, reducing the number of stages and
the number of devices per stage
Power transistors

The next decade promises to be equally exciting
GaN device technology is maturing
 The first devices have reached production status


Output power of 100 W per device at X band
frequency have been reported by laboratories


Currently 25 W from MESFET devices
The devices hold the promise that powerful
solid-state transmitters will become reality at
frequencies up to X-band in the next decade
PCB Materials

Rogers introduced its 4000 series of laminates in
the early 90’s.
This substrate material is compatible with fiber-glass
FR4 PCB and can be stacked up with fiber-glass to
make multilayer PCB’s handling signals up to mm
wave frequencies
 This material is steadily impacting the way
professional microwave systems are built


Slowly but surely extending RF construction techniques to
microwave and millimetre wave frequencies
Microwave systems

Traditionally, microwave components such as mixers
and amplifiers are individually packaged in
connectorised enclosures



These modules are expensive, with the enclosure contributing
most of the cost
At the same time they provide a very good shielding solution
to contain radiation from microwave components
With materials such as Rogers R4003, the PCB can
become a microwave enclosure at a much reduced cost
A PCB-housed 6-8 GHz doubler
Construction






Top layer: Rogers R4003
The cover is bent from thin sheet metal and soldered
into place as a shield
Bottom shielding is provided by ground plane layers
inside and on the surfaces of the PCB
DC and BIT circuitry is housed outside the RF shield
All connections into the RF volume is made through
LC feed-through components via an internal layer
The key to a successful design is to know something of
antennas…
Antennas 101
Take care!

Example of how not to launch (ca 2000)
Poor
Earthing
Gap!
Slot
Antenna
Oops!
Launchers

There is obviously a serious leakage problem at
the connector transitions


Totally unacceptable for use in a complex system
Launchers should be chosen with care
The transition must be modeled and simulated
Not only for S11 and S21
 Also for leakage!

PCB Design

The PCB electromagnetic problem is a 3D problem




Transitions must be carefully designed to prevent radiation
(or susceptibility to outside radiation)
The board layout must be carefully planned from a functional
and also from an EMC view
Look at the board from all sides!
Following is an example of a recent low-leakage (but
not leakage free) edge launcher design on an evaluation
board
Low Leakage Launcher
PCB design

PCB enclosures potentially offer big cost savings in
microwave circuitry




The PCB design problem is far from trivial.
Too complex and time-consuming to simulate EM fields fully,
forcing us to make judicious use of EM analysis tools
Knowledge of where current flows and good insight into the
behaviour of electromagnetic fields is essential for creating
good PCB designs
PCB enclosure design is challenging!

A source of interesting research questions for the academics!
Finally

The technological advances that had the greatest
impact on RF progress?
Undoubtedly the maturation of 3D EM analysis
packages
 Passive components such as filters and antennas
often need only one design iteration
 Without these tools, the sophistication reached with
RF active and passive components would be
unthinkable!
