Fourth Year Projects for 2001-2002


Group Project in CMOS Bluetooth Transceiver Design

What Actually Happened This Year

In this year, the project was a low IF blue tooth transceiver in 0.35 micron CMOS. Background research came first in which the students learned about the CMOS process, principles of low-IF down conversion, RF and analog design procedures and techniques. Then students designed their individual blocks, first reviewing what others (previous students at Carleton and designers elsewhere) then one or several designs were chosen and detailed design was done on their individual blocks. As well, attention was paid to how the blocks would connect to the neighboring blocks.

Students involved and the blocks they worked on are as follows:

Presentations Done

Part of this group presented their design at Conexant in Ottawa, and also at the MR&DCan Workshop (Microelectronics Research and Development in Canada). This presentation is posted next to the elevator on the fifth floor Mackenzie Building.

Circuit Fabrication

Part of the group had their design fabricated after the term was over with components coming back in the end of the summer. Components which were fabricated were the LNA, mixer, oscillator and power amplifier. The decision to fabricate or not was based on a number of factors: available fabrication space, the likelihood of success, the novelty of the design, willingness and availability to do more work (e.g., completion of layout and testing) after the term was over.

Information From Before the Term, actually from previous years


Plots of last year's RFIC Design

Students will design a radio transceiver in 0.35 micron CMOS. This will involve a significant amount of time to become familiar with RF circuits, the CMOS process and the simulation and layout tools. Students will work as a group, but will also be responsible for individual parts which make up the transceiver. Parts could include the low-noise amplifier, image-reject filter, mixer, oscillator, synthesizer, and power amplifier. If good progress is made on the radio, or if individual parts look very promising then actual fabrication through the Canadian Microelectronics Corporation will be considered towards the end of the school year.

This project assumes a good understanding of background theory as provided by the third year courses, Electronics II (97.359) and Communications (94.351). Further theory will be provided by some of the fourth-year courses.

More Details on Some of the Parts, more to be added later.

1. Radio Frequency Mixer Design

In a radio-frequency front end, the incoming signal at the radio frequency (RF) is converted to the intermdediate frequency (IF) by mixing with a local oscillator (LO) signal. Thus the mixer inputs are the RF and LO signals and the output is the IF signal. Typically, many RF signals may exist, but only one is desired. This means that one of the main concerns in a mixer is linearity of the RF input to prevent intermodulations between various input signals. Other issues of importance are frequency response, power dissipation and noise.

This project will begin with a study of different types of mixers. The study will involve simulation with a combination of HPEESOF and SPICE to evaluate the linearity, and other specifications of the different structures. Then a particular design will be chosen for detailed analysis and design, followed by implementation and test.

2. Radio Frequency Oscillator Design

Oscillators are used in several ways in Radio Frequency circuits, for example as an input to a mixer which converts from the radio frequency to the intermediate frequency. Usually, the oscillator frequency is adjustable (Voltage-Controlled Oscillator or VCO) and used inside of a frequency synthesizer. One of the biggest concerns in oscillators is noise components which can cause the wrong frequencies to be mixed down to the intermediate stage.

This project will begin with a study of the background material and of possible oscillator topologies. This will be followed by simulation, probably with HPEESOF, to determine which type of oscillator results in the lowest noise. Then, a particular topology would be chosen for more detailed study of the design tradeoffs, followed by implementation and test.

3. Radio Frequency Power Amplifier Design

Power amplifiers are used to drive the antenna on the transmit side. Difficulties in design are achieving high enough power while using a battery as a power supply. High efficiency is important in order to maximize life of the battery. Different classes of amplifiers (from class A through class F) can be chosen with different tradeoffs between efficiency, linearity and simplicity of design.

This project will begin with a study of the different classes and types of power amplifiers. This will be followed by simulation, probably with HPEESOF, to determine which type of amplfier results in the highest efficiency and what the tradeoffs are. Then, a particular design, or combination of designs will be chosen for detailed analyis and design, followed by implementation and test.

4. Radio Frequency LNA Design

The LNA provides amplification close to the input without adding too much noise. This means that noise added by later stages is of less importance. Thus the main issue is noise. Other things are also ofconcern. For example, the input impedance must be matched to the antenna. Linearity must be sufficiently high to avoid overloading in the presence of high amplitude blocking signals, while still receiving potentially low amplitude wanted signals. Other important design considerations are power dissipation, layout area etc.

This project will begin with a study of different types of LNAs. The study will involve simulation with a combination of HPEESOF and SPICE to evaluate the noise, linearity, and other specifications of the different structures. Then a particular design will be chosen for detailed analysis and design, followed by implementation and test.

4. Synthesizer Design

In a radio communication system, one can typically communicate on one of many channels. In a radio front end, the synthesizer is sets the local oscillator frequency to tune in the desired channel. Thus the synthesizer contains a voltage-controlled oscillator (which is a project of its own) but also adds the control circuitry to set the frequency and to allow selection of frequency. Synthesizers can be based on phase-locked loops, of which there are several variants, or can be direct digital (where the equivalent of sine waves are stored in memory and read out at an appropriate rate to produce the desired signal). Issues in synthesizers are switching speed, resolution, tuning range, etc.

This project will begin with a study of different types of LNAs. The study will involve simulation with a combination of HPEESOF and SPICE to evaluate the noise, linearity, and other specifications of the different structures. Then a particular design will be chosen for detailed analysis and design, followed by implementation and test.

Courses most strongly related to the project