Designing a Bluetooth Transceiver in RF CMOS

Document Sample
Designing a Bluetooth Transceiver in RF CMOS Powered By Docstoc
					            Case History: Development of a Bluetooth Transceiver in RF CMOS
                   By Bob Koupal & Marshall Wang, Signia Technologies
                           & Cory Edelman, Agilent Technologies

The design of Signia Technologies’ Ulysses SBT5010 2.4-GHz Bluetooth transceiver required
special attention to achieve its design using a low-cost CMOS RFIC process. Key performance
criteria were evaluated at all stages of the design process using EDA software that included
integrated foundry models, system-compliance test templates for Bluetooth test signals and links
to test instruments to verify that measured results agreed with simulation data.

The Bluetooth standard was established for short-range wireless transceivers that are embedded
in a wide range of devices, including notebook computers, printers, fax machines, cameras, cell-
phones, head-sets, home appliances, PDAs and other portable devices. Manufacturers of these
devices require that the addition of Bluetooth be accomplished with small size, low power
consumption and low cost. To meet this demand, a System-on-Chip (SoC) solution using a high
level of integration is required. The goal of SoC is to eliminate expensive external components
such as surface acoustic wave (SAW) filters and inductors, to minimize the number of external
passive components and reduce overall cost, eliminating external RF tuning elements to simplify

Interference is a significant design issue for Bluetooth radio links, which operate in the non-
licensed 2.4-GHz ISM band with a signal that is frequency hopped over eighty 1-MHz channels
at a rate of 1600 hops per second. Bluetooth links must have reliable operation in the presence of
other RF emitters in the 2.4-GHz band, including other Bluetooth devices, cordless telephones,
microwave ovens and wireless LAN systems such as those using the IEEE 802.11 standard.

A single-chip radio has been the stated goal for many wireless standards, including GSM and
CDMA phones, but Bluetooth is the first standard to truly demand it to be successfully
implemented in high volumes, with small size and low cost. Bluetooth IC designs require
increasing levels of complexity as radio, mixed-signal and digital functional blocks are combined
on a single piece of silicon.

The Signia Two-Chip Solution
Signia Technologies has implemented a Bluetooth SoC solution on two chips, a 2.4 GHz
transceiver RFIC (see Fig. 1) and a baseband controller IC. This approach optimizes the Bill of
Materials (BOM) costs, with the flexibility to address a number of applications. The radio IC
includes a cascode-type low-noise amplifier (LNA) with a source-degenerating inductor for
noise matching and a 50-Ω input impedance. The LNA has a 4-dB noise figure, 15-dB gain and a
-10 dBm input third-order intercept point (IIP3.) The receiver also includes an IF filter, mixer
and demodulator. The transmitter includes a mixer, filter and power amplifier (PA.) The PA is
matched to 50 Ω using an on-chip high-Q inductor and delivers 20-dB gain, +8 dBm P1dB, and a
power-added-efficiency of 15% at P1dB.
   Fig. 1: The Signia Ulysses SBT5010 Bluetooth TxRx showing the primary simulation
                               methods used for each block

The chip also includes a Gaussian Frequency-Shift Keyed (GFSK) modulator/demodulator, a
2.4-GHz PLL controlling a quadrature voltage-controlled-oscillator (VCO), tunable IF filters,
digital clock recovery circuitry, dc estimation and control logic. The VCO is implemented
completely on the RFIC with an on-chip resonator and frequency-tuning varactors. The VCO
consumes only 5 mA, and its single-sideband phase noise has 10 to 15 dB margin beyond the
level required by the Bluetooth specification. The baseband IC (not covered in this article)
contains the entire core Bluetooth functionality, with interfaces for USB, UART, and an
innovative host interface that allows the controller to be configured for PCMCIA, parallel/printer
port, and PCI interfaces without the need for additional components. The two chips communicate
through an 8-pin BlueRF mode-2 bi-directional interface.

Bluetooth RF Design Issues
Complexity, combined with cost-related goals, requires that the design of a Bluetooth radio SoC
take a different approach to simulation and layout. Traditional RF building-block design would
start with components, then develop a circuit topology, which is analyzed using harmonic
balance or circuit envelope simulation. This sequential design method does not work for SoC
design; instead, several disciplines are applied simultaneously, since portions of the circuit are
high-frequency analog, portions are mixed-signal, and the remainder are purely digital. Some
parts require linear, nonlinear or electromagnetic RF simulation, others require both SPICE-like
simulations and Verilog for timing analysis, while some require only Verilog simulation. In this
design, there are several hundred active components in the analog portion, and over 30,000 gates
on the digital side. The complexity of this design prevents the designer from using a single EDA
tool for simulating the entire chip: System-level planning must ensure that the individual blocks
are specified correctly and simulated using the right technologies.

Power and noise budgets are among the first, and most important, system parameters that define
performance. Bluetooth is nominally a 1-mW (0-dBm) system, using GFSK modulation in a 1-
MHz bandwidth. Together with a 0.1% Bit Error Rate (BER) requirement, this implies a Carrier-
to-Noise (C/N) ratio of 21 dB. The thermal noise floor for a 1-MHz bandwidth is -114 dBm, and
the Bluetooth specification demands a receiver sensitivity of -70 dBm or better, or a receiver
noise figure of 23 dB, which is based on the typical transmitted signal level at a range of 10 m.
However, realistic applications of Bluetooth require 10 to 15 dB better receiver sensitivity to
overcome the effects of the difficult indoor propagation environment (see Fig. 2.)

 Fig. 2: The signal and noise level relationships governing Bluetooth receiver performance

For the requirements of Bluetooth, there are several options for a process technology:
• Complementary Metal-Oxide Semiconductor (CMOS)
• Bipolar-CMOS (BiCMOS)
• Silicon-on-Insulator (SoI)
• Silicon-Germanium (SiGe)

Of these processes the lowest cost is CMOS. For example, the 0.25-µm CMOS process provided
by Taiwan Semiconductor Manufacturing Company Ltd. (TSMC) and other companies is
typically 30 to 50% lower in cost than BiCMOS because it requires fewer mask and process
steps. SoI and SiGe have more expensive starting materials and more complex processes. CMOS
achieves its low cost though high volume, and there are many foundries with available capacity.

To produce the RF blocks necessary for radio system design, TSMC’s well-established 0.25 µm
process provides an fT of more than 30 GHz, which enables the design of the required Bluetooth
building blocks – a 3.5-dB noise figure LNA with 20 dB of small signal gain, a fully integrated
VCO with quadrature output and phase noise better than –95 dBc/Hz at 500-kHz offset, and a
fully integrated 10-dBm PA with 30% efficiency. The process has been tailored for analog RF
design, with features that include thick top metal for fabrication of high-Q inductors and
transformers, metal-insulator-metal (MIM) capacitors, multiple VT devices and varactor diodes.
The process also has deep-n-well structures that improve device isolation.

The overall system architecture is the next step in the design phase. Beginning with the receiver,
we recall that Bluetooth specifications require at least a –70 dBm sensitivity with a BER of
0.1%, even with interference present, with –80 dBm or better needed for most applications. For
cost and manufacturability this performance must be achieved with a minimal number of
external components. There are 3 basic receiver architectures that might be considered, each with
its own advantages and limitations:

•   Traditional superhetrodyne (see Fig. 3) – The superhetrodyne topology easily meets the
    electrical requirements. They have high channel selectivity and are well understood circuits,
    but require external SAW filters making them too expensive for Bluetooth requirements.

                         Fig. 3: Superhetrodyne receiver architecture

•   Zero-IF (see Fig. 4) – Zero-IF designs allow high levels of integration and eliminate the need
    for an image filter, but have poor rejection of local oscillator (LO) feed-through and can have
    dc offsets on the outputs, making zero-IF a relatively complex design to implement.

                              Fig. 4: Zero-IF receiver architecture
•   Low-IF (see Fig. 5) – Low-IF was selected by Signia, a design architecture that also supports
    high levels of integration and eliminates the need for external filters, but is simpler to
    implement than the zero-IF approach, and leads to higher yields and lower cost.

                             Fig. 5: Low-IF receiver architecture

Once the low-IF receiver architecture was selected, the next decision was whether to use limiter-
based or I-Q demodulation. I-Q receivers are commonly used in digital communication circuits
for QPSK demodulation, but require accurate quadrature LO signals and a high-linearity AGC
amplifier. The limiter-based method is well suited for 2-FSK applications and, without needing
an AGC amplifier, is much simpler to implement, which lowers the cost. Another design choice
was a matched 50-Ω receiver front-end that did not require external matching networks. The
entire low-IF topology was verified in simulation and optimized for the desired performance
using high-level behavioral blocks included in the Bluetooth DesignGuide which is part of
Agilent EEsof EDA's Advanced Design System 2001 (ADS 2001.)

The Bluetooth transmitter specification permits three levels of output power: 0 dBm (Class 3),
+4 dBm (Class 2), and +20 dBm (Class 1.) Class 1 enables a range up to 150 meters, and its +20
dBm power level is usually going to be implemented with an external PA. In this design, care
was taken to keep the 2-MHz spur of the transmitter below –40 dBc to stay within specifications
if an external PA is added. To determine the optimum transmitter topology two approaches were
considered: Direct VCO modulation and I-Q modulation. The direct VCO approach is attractive
because it a simpler design and requires less current, but it is prone to problems with frequency
drift and variation of the modulation index. The approach selected for this design was I-Q
modulation, which effectively eliminates frequency drift problems and allows direct modulation
using I and Q signals derived from the baseband IC. As with the receiver, the entire transmit
topology was verified using Bluetooth DesignGuide in ADS.

Simulation And Verification Of The Designs
RF circuits designed to fit the selected topologies were initially simulated using the ADS
harmonic balance simulator, in conjunction with a Design Kit that includes RF models provided
by TSMC. The Design Kit has symbols created for each supported device type in the TSMC
library, including corner models. The devices include a variety of BSIM3 models for nMOS and
pMOS transistors, MIM capacitors, a number of polysilicon and diffused resistors, inductors,
varactors and other diodes.

To derive on-chip baluns or specific inductor values that are not supported by the Design Kit, the
devices are designed separately and the file then attached to a symbol for use in simulation. They
are first laid out using Cadence Analog Artist, then imported into ADS’s Momentum RF 2.5D
electromagnetic simulator. Pins are added to the layout to define the current flow direction, the
polygons are meshed into rectangles and triangles, and the dielectric properties of the substrate
are defined. The circuit is then simulated using the planar field solver, and the S-parameters are
created and stored in a model file.

When samples are received from the first tape-out they must be tested. If the only measurements
available this early in the process use simple sine-wave sources, only basic verification steps can
be performed – such as gain, output power and other frequency-domain measurements. While
this will indicate that the RFIC is somewhat functional, it does not provide confidence in its
performance for the desired application. To perform application-specific testing, there are a
number of dedicated signal sources that feature a Bluetooth personality, along with
complementary signal analyzers. This equipment should be linked into the simulation
environment to ensure that the same stimuli and measurement algorithms used in the design
phase will be available in the test lab.

          Fig. 6: Bluetooth signal source (ESG) and Vector Signal Analyzer (VSA)
           operated by the EDA-based DesignGuide to verify RFIC performance

Fig. 6 shows how EDA-linked test instruments allow the designer to connect to the EDA
software in a way to ensure that repeatable, realistic Bluetooth test signals and measurement
algorithms are used for simulation and hardware design. It is also possible to insert simulation
models into the signal path at any point. These models may be hardware, such as portions of the
circuit that are not part of the RFIC, or they may be signal impairments such as multipath fading
or interfering signals of any type. The Bluetooth 1.1 specification spells out what is required for
receiver performance: Sensitivity, adjacent-channel interference at various frequency offsets, co-
channel interference and intermodulation distortion. For each of these tests the system must
achieve a BER of 0.1% (see Fig. 7.) This is the so-called “raw” BER, which does not include any
improvements gained by error-correcting coding.

 Fig. 7: A critical test for the low-IF receiver’s performance is adjacent-channel selectivity
Verifying The Design

               Fig. 8: Receiver Sensitivity Template in the EDA environment

Fig. 8 shows a typical test template that combines both the EDA environment and test equipment
for hardware design and verification. The reference transmitter and/or receiver model is replaced
by the circuit under test, which is then compared with the simulated design until the hardware
becomes available. The signal path may include a simple Additive White Gaussian Noise
(AWGN), or Multipath model optimized for indoor propagation. Performance may be observed
by using an eye diagram (Fig. 9), a BER measurement (Fig. 10), or via a software-based Vector
Signal Analyzer to demodulate the signal and measure modulation accuracy (Fig. 11.)

               Fig. 9: Eye diagrams at the test receiver’s demodulated output
                   with a -70 dBm input (left) and -84.5 dBm input (right)
Fig. 10: BER vs. received signal power for the candidate receiver, which exceeds
      the -70 dBm Bluetooth requirement, and should be usable to -83 dBm
 Fig. 11: Vector Signal Analyzer showing the demodulated symbols vs. time, eye diagram,
           FSK error percent and received data statistics and the signal spectrum

As noted earlier, a variety of simulation technologies must be used for complete and accurate
verification. Initial analysis of circuit design can be started using non-linear steady-state
techniques such as harmonic balance. Using single or multiple tones and their harmonics,
harmonic balance is applicable to mixers, amplifiers and filters. For I-Q modulation with direct
up-conversion to the desired transmit frequency it is difficult, if not impossible, to infer the
quality of performance using tone excitation; the full I-Q representation of the modulated signal
must be used to drive the modulator. Although a transient simulation can infer some aspects of
modulator performance, only an circuit envelope simulator can predict it fully. Therefore, the
EDA environment must provide either the signal-processing necessary to produce the I-Q
representation of the Bluetooth signal and be able to apply the envelopes to the circuit being
evaluated, or that signal must be created ahead of time and stored in a format usable by the
circuit envelope simulator.

Finally, the schematic was transferred to a layout environment and subjected to time-domain
simulation as an additional verification step. Following the layout entry and design rule check, a
GDSII stream was produced and sent to the foundry. Finally, the fabricated devices were verified
in a full transmit or receive configuration, again using the EDA-based test templates and
associated test equipment.

Simulation And Verification Results
To review, the simulations were performed using both steady-state and time-domain methods.
Harmonic balance techniques were used for the initial design of the I-Q modulator, power
amplifier, LNA and mixer subcircuits. Two time-domain methods were used. First, SPICE was
used to examine transient performance where needed. The Circuit Envelope simulation method
was used to examine the time-domain co-simulation of complete subcircuits including signals
and measurements, operating in the Agilent Ptolemy-based behavioral signal-processing
environment. With the appropriate simulation tools available, there was no need to compromise
accuracy by introducing simplistic behavioral models for the RF subcircuits into the compliance
test templates.

A sample of measured and simulated results for the power amplifier subcircuit of the transceiver
are shown in Fig. 12. In this case, measurements of the first-pass design all comfortably
exceeded simulated performance parameters.
Fig. 12: Simulated and measured performance for PA (output power/gain vs. input power)

This article describes the process of definition, design, simulation and verification of Signia
Technologies’ Ulysses SBT5010 Bluetooth RF Transceiver, implemented using a 0.25 µm
CMOS process. The design process eases the rapid prototyping and design of an RFIC using an
EDA environment that includes several key features – availability of different nonlinear, time
domain, electromagnetic and signal processing simulation technologies, integrated foundry
models, compliance test templates, and links between simulation and test instruments.

To learn more about ADS 2001 or the Bluetooth DesignGuide visit

The authors wish to acknowledge the cooperation and support of Agilent Technologies, Inc. and
Taiwan Semiconductor Manufacturing Company Ltd.

[1] Bluetooth SIG web site
December 1999.

Shared By: