This group project was associated with our Power Electronics Laboratory where we designed and developed of a mini offline UPS system tailored for low-power devices such as routers and modems during power outages. The system aims to provide a 220V AC output at 50Hz frequency, capable of delivering up to 1A current. Key features include automatic switching, overcharge protection, and efficient power delivery. The design employs rechargeable lead-acid battery, voltage regulators, and an inverter circuit to convert stored DC energy into AC power. Given the tolerance of low-power devices to non-sinusoidal waveforms, the output is maintained as a square wave without additional filtering.
Find more about it here.
This is a group project associated with Digital Electronics Laboratory where we implemented Simple-As-Possible (SAP-1) computer on a CPLD using Verilog HDL. This implementation includes fundamental components such as a memory unit, a program counter, an instruction-register, a simple arithmetic logic unit (ALU), and control logic, which together execute a basic set of instructions. The project demonstrates the step-by-step process of building the SAP-1, from initial design specifications to hardware implementation, along with simulations and testing to validate the correct functionality of each component. The results provide insights into the working of basic computer architecture and lay the foundation for more advanced digital system designs.
Find more about it here.
This project is associated with Control Systems Laboratory where made a Self-Balancing Reaction Wheel that achieves stability by generating counter torques using Full State Feedback Control mechanism with a Linear Quadratic Regulator (LQR). The feedback mechanism utilizes the angular position, angular velocity of the body and the angular velocity of Brushless DC motor as state vectors that drive the voltage of the motor to attain stability. We took the data using a gyroscope and a magnetic encoder and implemented the control mechanism using an Arduino UNO and a motor driver to drive the DC motor for balancing.
Find more about it here.
This project is associated with Digitial Signal Processing Laboratory where we worked on an OCR system for Bangla word. The algorithm involves multiple stages like image preprocessing, character segmentation, character detection and word reconstruction. We used OpenCV and other relevant python libraries for this project and achieved accurate character segmentation using connected component analysis. For character detection we used a CNN network for multiclass classification where it detected 60 classes with 91% accuracy. Finally, we reconstructed the word from characters. The entire system was incorporated with a GUI made using the tkinter library.
Find more about it here.
This project is associated with Power Systems Laboratory where we did a modelling and simulation of a three-phase grid-feeding inverter using MATLAB Simulink and analyze its performance for connecting a solar PV panel to the IEEE 5 bus system. Using MATLAB Simulink, a model of a grid feeding inverter is implemented and simulated under various operating conditions. The simulation results are analyzed to evaluate the performance of the inverter system in terms of power quality, efficiency, and grid synchronization. Additionally, the impact of control parameters and system configurations on the inverter's behavior is investigated to optimize its performance.
Find more about it here.
The project is about implementing a frequency demodulation circuit where we address the problem of retrieving a message from an FM carrier signal. FM signal demodulation presents challenges due to its non-linear relationship between the input and output signals, making the recovery of the original message signal more complex compared to Amplitude Modulation (AM). Primarily, we looked for a simple and implementable method for demodulation. After multiple considerations, we chose Balanced Slope Detector circuit the most suitable for implementation. Throughout this project we design, simulate and validate the performance of our designed balanced slope detector circuit.
Find more about it here.