By: Roghoyeh Salmeh

1

Direct Conversion Modulation/ De-Modulation at
Microwave Frequencies

• By Roghoyeh Salmeh
• Supervisors Dr. Brent Maundy and Dr. Ronald
  Johnston

1
2
DIRECT CONVERSION MODULATION/DE-MODULATION AT
MICROWAVE FREQUENCIES
2
3
INTRODUCTION

• Portable equipment such as cellular phones and
  wireless local area network (LANs) are in strong
  demand.
• One effective way of decreasing the power
  consumption, weight and size of a transceiver is
  to use simpler circuitry with lower power supply
  voltage.
• Improving circuit techniques and transistor
  technology scaling will contribute advances
  toward this goal. Architectural innovations in
  the transceiver may lead to revolutionary
  improvements.

3
4
INTRODUCTION (Cont.)

• For Microwave telecommunications, increasing the
  spectral efficiency or transmission capacity is
  also very important.
• Usually increasing spectral capacity is obtained
  by sophisticated amplitude and phase modulation
  techniques such as QPSK.
• The baseband section of a transceiver has more
  components and is more complex.
• The digital part of a mobile transceiver is
  already highly integrated but in the analog RF
  part a higher integration level is still strongly
  needed.

4
5
MODULATION ARCHITECTURES

• There are two major architectures to perform
  modulation/de-modulation
• 1) Super-Heterodyne architecture
• 2) Direct Conversion architecture
• Super-Heterodyne modulation technique was
  introduced by Armstrong in 1918 and was generally
  thought to be the receiver of the choice due to
  its high selectivity and sensitivity.
• In this architecture the signal band is
  translated to a much lower frequency called the
  intermediate frequency (IF).

5
6
SUPTER-HETERODYNE RECEIVER ARCHITECTURE

• In super-heterodyne designs, the input RF signal
  is first amplified and then converted to a lower
  intermediate frequency (IF) and then converted to
  the base-band.
• Today almost 98 of the radio frequency receivers
  use this architecture.
• This architecture has limitations for achieving
  miniature and low power transceivers.

6
7
SUPTER-HETERODYNE ARCHITECTURE (Cont')

• Since in this architecture the translation from
  RF to baseband happens in a series of steps, a
  receiver of this type needs at least two local
  oscillators, two mixer stages and filters. This
  in turn will increase the chip area and power
  consumption.

Figure 2. The block diagram of a super-heterodyne
receiver with one IF stage.
8
SUPTER-HETERODYNE ARCHITECTURE (Cont')

• The RF signals in this architecture must be
  routed off-chip to an image frequency filter.
  These image reject filters typically are not
  built on-chip.
• Off-chip routing of the RF signal requires that
  the output of the amplifier to be matched to 50
  ohm. This in turn reduces the amplifier's gain.
• Amplification and filtering at IF frequencies
  come at the price of power dissipation and the
  need for off-chip passive components.

9
DIRECT CONVERSION MODULATION/ DE-MODULATION
ARCHITECTURE

• Direct conversion, also called Zero-IF was
  introduced many decades ago and is the natural
  approach to downconvert a signal from RF to
  baseband.
• Since its inception, it had been tried many times
  but had not achieved sufficient performance to be
  widely used.
• Direct conversion is presently limited to only a
  few applications such as pager receivers.

10
DIRECT CONVERSION MODULATION/DE-MODULATION (Cont')
Direct conversion has recently become the topic
of active research to a much greater extent than
before. The main reasons for this renewed
interest are 1) Direct conversion allows
monolithic integration more easily. 2) Direct
conversion suffers less from mismatches caused
effects. 3) The problems that appeared in
discrete implementation of direct conversion may
be controlled and suppressed in IC implementation.
10
11
DIRECT CONVERSION MODULATION

• The major advantages of direct conversion over
  the heterodyne system are
• Simplification
• Performance

11
12
DIRECT CONVERSION DESIGN ISSUES

• RF circuits must process analog signals with a
  wide dynamic range at high frequencies.
• The signal in RFIC design are treated as analog
  even if the modulation is digital.
• There is always trade-offs in RFIC design, which
  can be summarized in RF design hexagon.
• This hexagon includes Power, Frequency, Gain,
  Supply voltage, Linearity and Noise.
• Digital circuits directly benefit from advances
  in IC technology but RF circuits do not as much.

12
13
DIRECT CONVERSION DESIGN ISSUES (Cont')

• Direct translation from microwave frequency to
  base-band (zero frequency) entails a number of
  issues that do not exist or are not as serious in
  a heterodyne system. Some of these issues are
• DC offsets
• I/Q mismatch
• Flicker noise
• Even order distortion

13
14
DC OFFSETS

• DC offsets are considered the most serious
  problem of direct conversion.
• DC offsets arises because a small fraction of
  the local oscillator energy can leak through the
  mixer and LNA (due to their finite reverse
  isolation) and then be self mixed in the mixer to
  produce a DC offset.
• This offset appears in the middle of the
  downconverted signal spectrum and may be larger
  than thermal and flicker noise.

15
I/Q MISMATCH

• A direct down/up conversion architecture must
  comprise quadrature conversion.
• Creating the I and Q components from the RF
  signal is not easy and can cause severe
  noise-power gain trade-offs
• The amplitude of the I and Q components at the
  output of the LO must be tightly matched because
  any mismatch of I and Q signals corrupts the
  downconverted signal.

16
FLICKER NOISE

• The flicker noise problem arises due to small
  amounts of RF amplification in the receive chain.
• The coupling of the flicker noise with the
  received signal occurs primarily after down
  conversion to the baseband.
• Flicker noise is present in all semiconductor
  devices and its frequency depends on the
  processing technology.

17
EVEN ORDER DISTORTION

• Non-ideal behaviour in all components of Fig. 3
  can cause distortion in the RF signal.
• Although all RF receivers are affected by
  odd-order intermodulation effects, direct
  conversion receivers must also have sensitivity
  to even-order distortion.
• Even order distortion effects can be minimized
  through the use of differential circuitry.

18
FABRICATION TECHNOLOGY

• A proper fabrication technology can relax some
  of these issues.
• The fabrication technology must satisfy a low
  noise requirement and maximum frequency of
  operation.
• The ease of digital integration is also very
  important because it allows fabrication of the
  receiver as a System-on-Chip (SoC) with
  integrated digital parts.

19
RESEARCH OBJECTIVES

• The focus of this research is to design an
  analog direct conversion receiver at microwave
  frequencies for 3rd generation of wireless
  communication systems.
• Work to extend the range of frequency of
  operation and lowering the power supply for
  direct conversion receivers.
• Investigation into possible new VLSI
  architectures for direct conversion receivers.
  Implementation of the proposed architecture into
  integrated circuit.

20
PRESENT STATE OF THE WORK

• Specifications such as gain and signal to noise
  ratio (SNR) of different building blocks of the
  receiver based on the IEEE 802.11a standard have
  been defined.
• An active direct down conversion mixer working
  at 5 GHz has been implemented in IBM SiGe
  technology and is awaiting for testing.
• A fully differential low noise amplifier based
  on bipolar transistors has also been designed.
  The LNA works at 5.3 GHz from a 2 V power supply.

21
5GHZ DIRECT DOWN-CONVERSION MIXER

• Most of todays double balanced mixers are based
  on the Gilbert cell. The cell, originally
  invented by Gilbert in 1967 is essentially an
  active mixer employing bipolar transistors.
  Gilbert cell has several attractive features such
  as
• Conversion gain
• Good isolation
• Integration capability
• Requires low LO power
• Fully differential structure.

Figure 4. The block diagram of mixer.
22
MIXER (Cont')

• The design of a mixer faces many compromises
  between conversion gain, local oscillator power,
  linearity, noise figure, port-to-port isolation,
  voltage supply and current consumption.

• The mixer in a direct conversion system must be
  more linear to attain the same performance as in
  the heterodyne system.

Figure 5. The layout diagram of the mixer.
23
CONCLUSION

• The direct conversion receiver eliminates many
  off-chip and on-chip components and offers
  significant power saving by amplifying a received
  signal mostly at DC rather than at an IF of tens
  or even hundreds of MHz.
• A proper choice of fabrication technology can
  relax some of the limitations of direct
  conversion receivers.
• Circuit and architectural innovations with
  proper choice of fabrication technology may lead
  to a revolutionary improvement in WLAN systems.

23