This week, we mainly worked on the final presentation and preparing for the final demo. Once presentations were over, I spent my time repairing small things on the board, helping with some integration testing, and working on slides for the public demo.
We are on schedule given that the demo is on Monday. In the next week, I plan to work with the rest of the team on finishing up the final report and demo video.
This week, we worked on full integration of everything from the hardware through to the software and the website.
The most significant risks to the project at this point is that something in the board hardware becomes damaged. We purchased spare components, so barring a few pieces, we would be able to swap in a new component. Another risk is magnets damaging pieces. While this is extremely unlikely, the solution to this would be 3D printing a replacement. This could be done very quickly.
No real changes were made to the design and we have not rescoped the project at all, so our schedule remains the same.
This week, we all worked on full scale integration with the hardware and the Raspberry Pi. This involved making sure the board could detect every single piece and could accurately determine the legality of a move and when a game should end. We initially had success, but later had some issues with the ADC on Wednesday. On Saturday, I was able to solve the problem by changing the way the “chip select” line was toggled in our setup. At this point, the process of piece detection, move detection, and legality checking works to our expectations.
In terms of the board mechanical aspects, our remaining magnets arrived this week, so we were able to finish assembly on the last pieces and test the board with every single piece on the board. I was also able to set up the Pi to have the whole system powered off of the board’s battery pack instead of needing an outlet for the Pi.
Given all of this progress, we were able to test both accuracy and power consumption with the expected load. The board was able to detect all moves accurately, and the power consumption is in line with what I predicted based on the Pi’s theoretical power use.
With this progress, we used some of the slack time, but have largely finished the final integration portion of the project. Any remaining small details will be ironed out next week before the demo.
In the next week, I will re-glue some of the pieces to make them more secure. I will also add some tape or latex to the bottom of each piece to increase the friction between the pieces and the board. This will prevent any pieces from interacting due to magnetic forces between themselves.
This week, I worked on finishing up the mechanical parts of the board. I laser cut some acrylic and engraved it with a chess board pattern that matches out sensor layout on the PCB. Also as part of the board mechanical work, I modeled a very simple magnet holder for the bottom of the pieces.
As part of testing, I tried to quantify the power use of the board. We know the rough power usage of the Raspberry Pi, so in order to verify that our battery capacity is long enough for 10 hours of use, we needed to know the power use of all components aside from the Pi. To last 10 hours, the entire system will need a little over 10.5 Ah of capacity. Our board can fit up to 24 Ah of batteries, so we have more than enough battery capacity.
Progress is still a bit behind in terms of the testing plan, but some tests like accuracy and power have been done. Many of the other tests rely on software integration with the firmware, and since these aspects of the project are behind, the testing is also behind. We will be using our slack time in the next week or two to finalize everything. We also will parallelize as much as possible.
For the next week, I hope to have all the piece and magnet assemblies done as well as have the acrylic attached to the board.
This week, we continued preparing for the interim demo. This included preparing some chess pieces for the board to validate piece detection. After the demo, I worked on preparing the acrylic board top by purchasing new acrylic and starting the design to be engraved on the panel. I also looked into buying a new set of chess pieces that allow a magnet to be more easily inserted into the base.
Progress for the hardware validation and testing with a microcontroller is on schedule in that this section of the project is complete and the system has passed our accuracy tests for piece detection. The overall progress for testing and integrating the hardware through to the website and testing the full-scale project is in progress, but about a week behind. To remedy this we will be trying to integrate performance tests during the integration. The latency and power use metrics can be done during this process. We also have two weeks of slack time to work with if necessary.
For the next week, I hope to have all the pieces ready to go with magnets in them as well as to have the acrylic top engraved.
This week, we spent the entire time preparing the hardware, firmware, and website for the interim demo.
The most significant risks to the project as of now are that integration takes too long and that we cannot meet a certain goal we set for ourselves. The issue of time is managed by the fact that we have two weeks of planned slack time. We also can run tests like power consumption while the board is not perfectly accurate.
As of now, no changes have been made to our requirements, our design, or our schedule. If we experience a few days of delay in progress, then we will likely end up using some of our two weeks of slack time.
Below is a picture of the computer correctly seeing pieces on the board (N is no piece, B is a black piece, and W is a white piece).
This week, we worked on preparing for the interim demo. For my part of the project, the interim demo involves getting a correct board state (piece color and position) using the sensors and the firmware. Patrick and I spent many hours debugging and testing the system. The current state is that the board can detect most squares correctly, but there are still a few that are incorrect. We will be debugging the rest of this tomorrow in preparation for the demo.
In terms of progress, hardware progress is on the edge of being a bit late, but assuming we finish debugging the last few issues tomorrow, we should be back on track. If debugging extends, we will be pushing into a bit of testing and slack time. However, the debugging tests do involve some of the tests for accuracy, precision, and power use. Therefore, we may be ticking off other items at the same time.
In the next week, I hope to fully integrate the hardware with the firmware and software.
This week, I finished designing and printing all of the threaded board standoffs and used them to assemble the full chess board from the 4 segments along with pieces of FR4. Once the board was fully assembled, we soldered all of the jumper resistors to make all the electrical connections between the 4 board segments. After doing this, I tested the sensor outputs to ensure all 64 sensors were making good contact with all pins. We also tested power regulation and the LED control.
Progress is still on schedule as we will begin the integration process between hardware and firmware this coming week in preparation for the interim demo.
In the next week, I hope to have the board in a state where the computer can sample a board state using all of the sensors. This means Patrick and I will need to work together on integrating the firmware with the ADC and multiplexers.
In the past two weeks, I ordered and received the PCBs and all of the components. I have also assembled all of the board and done some initial hardware testing to make sure there are no shorts anywhere on the board. Other tests I have run so far are for the power regulation, user interface (buttons, LED control), and the sensors to ensure they are all functioning correctly. The remaining tests are to be done with the multiplexers and ADC using an Arduino or the Raspberry Pi.
The other main part I’ve worked on is the chess board mechanical tasks. I designed and 3D printed standoffs for the board. I also cut some FR4 scrap to glue on the PCBs for support when linking them together.
I am still on schedule since the board is nearly ready for the beginning of integration work with firmware and the bulk of the mechanical aspects of the board are done.
In the next week, I plan to finish the mechanical aspects of the project and begin testing of the multiplexer and ADC interface with firmware.
This week, after doing the design presentation, I spent nearly all my time working on the layout and routing for the PCB. Throughout this process, I had to do some reworking of the schematic, but this was relatively minor compared to the other board work. The board is completely routed and ready to order. I will be doing some final checks and will review the entire board with the rest of the group.
My progress is still on schedule, but the ordering and shipping time may cause the board to arrive later than expected. This means that I may not receive the board and BOM until fall break or at the beginning of the week after fall break. This can be mitigated by making the mechanical board this week and during fall break since I planned to do that after break.
For the next week, I hope to have most, if not all, of the mechanical aspects of the board finished. This includes 3D printed standoffs and the laser cut board top.
Recent Comments