Auto-Guidenced HoverCraft (University Final year thesis project)

Group of 3 people had to design a hover craft prototype that had an Auto-Gudience System implemented. System will control the vehicle to go down an 18 meters hallway and stop automatically in the targeted area.

Report(1.6Mb) Picture 1 Picture 2 Picture 3
In ACTION(620kb)


Research Paper: Nanoelectronics

A research paper titled "The Development, Present Status, and Conceivable Evolutions of Nanoelectronics in Nanotechnology". Mainly discuss the evoluations of Nanoelectronics

Experiment report: Laser Doppler Velocimetry

Using Laser light to measure the speed of fluid. As well as determine the velocity profile of the flow.
Very interesting and useful method.
Main idea is to use the fluid flow to introduce a "phase shift" to the laser beams. As a result, doppler effect occur.

Abstract: Laser Doppler Velocimetry (LDV) can be used in measurement of flow rate as well as velocity profile in transparent flow channel. Relation between the frequency shift (f) and the velocity (v) was f=0.3245v. The sine of the beam angle is also directly proportional to the frequency. In addition to this, this experiment had also proved that LDV could use to determine the sizes and densities of particles in flow channel.


Research Paper: Hydrogen Fuel Cells

A research paper titled "Hydrogen Fuel Cell: The Development, Present Status and Conceivable Evolution of Alternative Automobile Engine". Mainly discuss the Development of Hydrogen Fuel Cells.


Experiment report: Thin Film Measurement

Unfortunately, this is an incomplete experiment.
The experiment was plan to use different technique to measure the thickness of thin film, including Michelson interferometry, Ellipsometry, and Fizeau. However due to time and limitation of equipments, result of Michelson interferometry was unavailable.

Abstract: The most efficient way to measure thickness of a semi-conductor sample and it's insulation layer, is to use Fizeau technique and Ellipsometry; Michelson provide more accurate value, but is more complicated


Independent Research: Feedback Temperature Controller

A Proportional Intergrational Differential (PID) Routine written in Turbo C that control a peltier device. User can set desire temperature, and the system will hold that temperature automatically. If the system heat up, the peltier will cool down by reverse bias the current. On the other hand, if the system is too cold, system will pump current in to heat it up.

A peltier can be see as a semi-conductor heat pump. More info can be found here.

Of course, this is not quite useful because it require a 286 machine (minimum) to run. However, the routine can be rewritten in Assembly language, and implement on a microcontroller.

This can be use as a electronic cooling device, maybe use it as a CPU temperature controller.

Next plan, re-write it in Assembly and have a PIC to control the peltier

C-Source Lab manual(PDF)

Independent Research: Digital Signal Processing, realtime decoding satellite image

A C language routine that will decode AM modulated weather satellite signal.

The original signal from the satellite was FM modulated. Our lab tech (Thanks Glen) had recieved and demodulated with an home-made antenna and a modified old tube radio. Then record the FM demodulated signal in to a wave file.

This signal is still AM modulated, and our job is to write a simple C routine that runs on a pentium 200MMX with Linux Mandrake as the OS to demodulate the signal in real time (DSP routine must excute faster than the stream), and re-generate an image.

Coding was very successful. All it takes to demodulate is 45 seconds on the 200MMX machine. The image contain 2 channels, one in visible spectrum and one in IR (infared). And we decided to use the IR channel to give colour to the picture. This make the code a bit slower, takes total of 80-85 seconds to complete the task. But it still way faster than then real time.

C-Source(B/W) C-Source(Color)Image(B/W) Image(Color) Lab manual(PDF)