Gene Chang’s Status Report for April 18

This week I focused on scaling and stabilizing the hardware subsystem for the LumiKey project, with the goal of transitioning from a small prototype to a full single-octave implementation that is reliable enough for validation and demo use.

Learning New Tools and Knowledge Acquisition

Since I worked primarily on the hardware components of the project, I had to learn how to interpret and apply technical documentation effectively. This included understanding the behavior of components such as FSR strips and LED drivers, as well as implementation details like bit shifting in the Adafruit LED libraries. To navigate this, I relied heavily on a combination of official documentation, online forums such as Reddit, and LLM based tools. These resources were especially helpful in quickly identifying relevant sections of large documentation sets, allowing me to focus on the specific details I needed rather than spending excessive time searching.

In addition to online resources, I adopted hands-on and collaborative learning strategies. I initially lacked confidence in soldering on a protoboard and troubleshooting hardware issues, but by consulting with TechSpark staff and seeking in person guidance, I was able to learn proper soldering techniques and best practices. This significantly reduced hardware related issues and improved the reliability of my system.

Through this process, I also developed practical debugging skills. For example, I was able to identify and resolve issues such as unintended short circuits caused by stray wires, which led to incorrect LED behavior during testing. Overall, combining targeted use of documentation, community resources, LLM assisted search, and hands-on mentorship allowed me to efficiently acquire the knowledge needed to design, implement, and debug the system.

System Expansion

A major milestone this week was expanding both sensing and visual output coverage:

  • Increased FSR input channels from 12 to 24, enabling support for a full octave range
  • Expanded LED output from 5 LEDs to 36 LEDs, allowing per-key visual guidance across the entire octave
  • Ensured proper mapping between each FSR input and its corresponding LED output

This expansion moves the system closer to the intended use case of guiding real piano playing rather than demonstrating isolated functionality.

FSR Strip Validation and Debugging

A significant portion of the week was spent verifying the reliability of the FSR sensing system:

  • Tested each FSR channel individually to ensure consistent and correct readings
  • Identified and debugged issues related to inconsistent triggering and cross-channel interference
  • Verified that each sensor responds appropriately to key presses and releases

This process was important to ensure that the system produces stable and interpretable input signals before integrating higher-level logic.

FSR Modification Experiments

I also explored the flexibility of the sensing hardware by experimenting with modifying the FSR strips:

  • Tested whether FSR strips could be cut to custom lengths for better physical alignment with keys
  • Found that the strips remain functional when cut within certain threshold regions, without breaking conductivity
  • This result improves the adaptability of the system for different keyboard sizes and layouts

Hardware Integration and Wiring

To improve system robustness and prepare for enclosure, I worked on consolidating the hardware:

  • Performed extensive wire management to reduce clutter and improve signal reliability
  • Migrated connections onto a perfboard-based PCB, replacing loose wiring with a more stable structure
  • Organized power, ground, and signal routing to support both LEDs and FSR inputs cleanly

This significantly improves durability and makes the system more suitable for packaging into a final enclosure.

Next Steps
Finalize enclosure design and integrate all components into a compact housing

Continue validating sensor reliability under continuous use
Integrate hardware with software pipeline for full system testing (input → processing → LED feedback)

Measure and report system-level latency and responsiveness

Leave a Reply

Your email address will not be published. Required fields are marked *