Digitaalisuunnittelu-kurssin tietokoneharjoitusten uudistus ja arviointi

This Master's thesis is concerns digital design course that teaches the basics of digital logic. Students on course learn to design, analyze and implement combinatorial gate networks, sequential systems and state machines on both theoretical level and on practical implementations with FPGA development board. The FPGA board that was previously used is aged and the computer exercises themselves have been fragmented between the greater course project and supplementary tasks. In this thesis, a totally new computer exercise project is designed. Before the development of the exercises begun, the new FPGA development board had already chosen. Selection was done based on the flexibility of the system. The new board can be used by wide selection of courses to create various types of tasks. Along with the new exercise project, a MOOC-based web portal, Plussa, was taken into use. This system allows a semi-automatic grading of exercise tasks. The objective of computer exercises is to combine the theory of digital design concepts and the practice done on paper exercises to work flow that includes real industry grade development tools. The subject of the exercise project is a game implemented on a LED matrix extension of the development board. A reference project was designed and based on that exercise tasks were created. Exercises were divided into five separate exercise sets: Introduction to tools and development flow, arithmetic and hierarchical design, state machine design, extending existing designs and integration. On return, tasks will be graded by an automated system that fetches the students implementation, uses logic simulator on the server to verify the correct functionality and grades the return accordingly. Without an automated grading system the grading would need to be screened manually from students screenshots. Hence, with automated system, the time spent by staff could be used more efficiently to benefit the course attendees. For students, the automatic grading gives instant feedback on returns. On average, students spent 54 hours on computer exercises, which fits well to the total time allocation of the course. Students found the exercises laborious but educational. Perception of the workload varied widely based on the background of the students. Intensity of the load could be lightened by improving instructions and by giving some of the designs as partially ready made. Alternatively, a graduated exercise model could be used for the exercises