MEng/BEng projects possibilities

Dr Robert Watson, Department of Electronic & Electrical Engineering, University of Bath
Robert Watson

Final year projects

This page just serves as a list of ideas for possible projects that I am willing to run. A lot of these involve both hardware (RF and general electronics) and software (Matlab, Python, C, C++).

  1. Most modern surface mount printed-circuit-boards (PCBs) are soldered using a technique known as reflow soldering. This is basically an oven into which you place the PCB with the components in place and "cook" solder paste until it melts and bonds the components of the PCB. Solder paste is basically small balls of solder suspended in a liquid flux which cleans the component and the board. When heated the flux is activated meaning it cleans the joint and largely vapourises. The solder balls then melt together to form a liquid which solidifies when the heat is removed and the joint cools. (see short video). To properly solder the components without damaging the PCB or components from thermal stress and overheating requires a carefully controlled temperature profile (see example from Actel). The aim of this project is to construct and evaluate a small, low-cost reflow oven. The heating elements could be quartz or ceramic infra-red devices (e.g., Ceramicx) controlled by zero-crossing solid-state-relay devices (e.g., Crydom). The temperature could be measured using a thermocouple device such as a MAX6675 and the whole thing controlled by an Arduino, PIC or other microcontroller. There are elements of danger here since the project involves mains voltage electricity and high temperatures meaning that close supervision will be required. Although we have a couple of reflow ovens in the new workshop this is still a worthy project.

  2. Time-Interval-Error measurement system. In order to quantify the performance of synchronisation clocks in telecommunications networks the time-interval-error (TIE) metric is often used. This is the time difference between the clock under test and a known reference clock. In telecommunications timing systems the reference clock is often an atomic standard such as a cesium oscillator. Loss of synchronization in telecommunications systems can lead to dropped calls in cellular system or complete paralysis of a network. The aim of this project is to develop a high precision (better than 1ns) TIE measurement system (e.g., based on Acam TDCs). The device will be capable of logging and analysing the data to calculate various measures of time stability including TDEV (time standard deviation) and MTIE (maximum time interval error). This project will involve both hardware and software. Measurement will be done against rubidium based atomic clocks in the department and possibly against a cesium and hydrogen-maser based clocks owned by collaborators (e.g., Chronos and NPL).

  3. Weather radar propagation modelling. The Met Office weather radar network in the UK operates at what is known as C-band which is 5.60-5.65 GHz. Weather radar receivers are really sensitive. They can receive signals as tiny as -120 dBm. The problem is that this also makes them really vulnerable to interference from other things in the same band such as the IEEE 802.11a wireless LAN. The propagation of the radar signals and the vulnerability to interference is strongly affected by the terrain and buildings. The aim of this project is to assess the interference problem using radio propagation models. The project will make use of (existing) modelling technique known as the parabolic equation model or PEM. Terrain data for the UK will be taken from the NASA SRTM (Space Shuttle Radar Topography Mission) data archive.

  4. C-band 5.60-5.65 GHz spectrum monitor. The objective of this project is to build an inexpensive, small spectrum analyser sensor with the aim of quantifying spectrum occupancy. The basis of the spectrum monitor is a very sensitive receiver electronically tunable over the 5.60-5.65 GHz band designed to identify and characterise IEEE 802.11a wireless LAN and other radio emissions that may cause interference to weather radar. Don't be put off by the RF nature of the project - the electronics required is fairly straightforward.

  5. Parabolic Equation Model acceleration using CUDA. The weather radar propagation modelling project will involve running a software model called the parabolic equation model or PEM. At the heart of the PEM computation is the fast fourier transform, A typical PEM calculation for a radar application needs to run many thousands (or even millions) of FFTs. Even on modern PCs this can take a considerable time to run. One possible solution to speeding up these calculations is to make use of the considerable computational power now offered by modern graphics cards. The nVidia CUDA software environment provides a means to access this computational resource in a relatively simple way. The aim of this project is to investigate the use of graphics cards to accelerate the PEM calculation. There are various possibilities for achieving this, ranging from just performing the FFT calculations on the graphics card with the remainder on the PC, through to performing the entire PEM calculation on the graphics card.

  6. Deployable Cubesat antenna systems. Deployable antennas are being considered for many new spaceborne radar missions. The humble umbrella is a simple example of a deployable structure. The basic concept is to design a collapsible structure that occupies little volume while in storage (in the nose fairing of a rocket) but that when deployed becomes a large and rigid structure which forms, for example, a reflector antenna. The quest for ever higher resolution radar imaging of the Earth's surface (and other planets) has requires ever larger aperture antennas. The objective of this project is to explore some of the fundamental problems of deployable structures. In the first instance we might want to try making a 3U-sized cubesat (The UKube-1 payload, TOPCAT, was built in the Department.) L-band antenna (1.42GHz).  This project is probably best suited to IMEE students but could be tackled by other students with an interest in materials and mechatronics.

  7. Development of an low-cost Vector Network Analyser. A vector network analyser is an RF tool that is used to measure the magnitude and phase response of RF electronics such as amplifiers. It can also measure how well matched the devices are at the input and output. Such instruments are typically two-port devices for measuring things amplifiers but can have up to eight ports. Normally these instruments are very expensive. A relatively basic instrument working up to 6GHz can cost around £65,000. The objective of this project is to create a low-cost version that works up to 1-2 GHz using low-cost software radio system costing around £2,000.

  8. GNSS-R antenna and receiver system for UAV

These are the first few that come to mind, but I have lots of other ideas. If you are interested in any of these, send me an email.


Group design and business projects:

Again just some ideas. Some aspects of these projects could possibly run on a reduced scale for individual final-year projects.

  1. Development of a soil-moisture radiometer payload for a UAV. Measurement of soil moisture is an important parameter used by soil scientists, meteorologists and ecologists. Since the amount of water in soil is related to plant growth and crop yield, the information is useful for agricultural applications and managing water resources. The European Space Agency has launched a satellite called SMOS to measure soil moisture, but this is relatively low resolution and the revisit time over the same area is low. The objective of this project is to develop a small, light-weight (ideally <1kg) soil-moisture radiometer that can be flown on the Department fixed-wing UAV or a small quad/hexacopter.

  2. Design of an ultra-compact FMCW synthetic aperture imaging radar (SAR) for a lightweight uninhabited aerial vehicle (UAV), such as the Department owned on. Even without flying the UAV testing could be done from the door window of a car/train! In addition to the main RF electronics in the radar payload there are a number of sub assemblies that could be considered: