Projects Years: 2023-2024
Contributors: Matthew Thomson, Kevin Ma, Nathan Wong, Omar Sheikh, Julia Hladky, Safwan Rahman, Michael Pacak, Mattias Ow
Altium365 Project: https://americas-starr-university-of-alberta.365.altium.com/designs/4B610E1A-5272-4254-9F6E-A4370CCA8F52
Github Repo: UASTARR/flight-computer-pcb
The SRAD Altimeter was STARR’s first attempt at designing a custom flight computer. Inspired mainly by the Raven 4 and Blue Raven altimeters, the SRAD altimeters used many of the same sensors to record flight data in an embedded Quad SPI flash chip, similar to the way a Blue Raven stores its flight data. It featured four independent e-match channels with continuity detection, a USB C port for data recovery, and an STM32L433RCT6 microcontroller to run the whole operation.
Due to a major design flaw in the switched-mode power supply, which snuck its way in due to a version control error and failure to notice the change, the SRAD Altimeter could not be powered by a LiPo battery as planned and we did not have time to fix the issue. The SMPS we chose was actually not a buck-boost converter as we thought, but a boost-passthrough converter. If the input voltage was below 3.3V, it would output 3.3V. If the input voltage was above 3.3V, the converter just passed the input voltage through to the output. This error should have been caught in a design review of the board, but we did not have any.
That meant when we naively supplied the board with a fully charged LiPo battery (~4.2V), many, but not all, components were immediately fried by exceeding their maximum rated voltage. The microcontroller survived and was debuggable, but it became clear certain peripherals and components were either not responding or giving garbage data.
Unfortunately, the switched-mode IC was very small and a specific Texas Instrument footprint, meaning replacing the chip was off the table considering the tools we had. Additionally, Launch Canada 2024 was rapidly approaching and we could not get components in time from Digikey. We also would not have had time to program the board. The firing channels, sensors, and flash storage were never fully tested. The screw terminals were way too large for the board. Because we had the PCB assembled through JLCPCB, we opted for smaller components which ultimately made the board essentially unworkable by hand.
Finally, this project was a case of too many cooks in the kitchen. We tried separating and delegating portions of the design to various members to make sure each part of the design would have dedicated focus to minimize mistakes. However, most engineering students unsurprisingly don’t really care when they have just a tiny cog in the machine to perfect. Experience has shown avionics members shine when they have a complex project they can really dive into and be proud of.
While the design for the most part was sound, this project highlighted the need for in depth review by multiple experienced eyes. Designing an altimeter is not a particularly difficult task, but doing so without making mistakes is. Programming and testing is hard and time consuming, so any time lost fixing design mistakes cuts into the time spent making the board actually function. We would fall into this pitfall next year. In order to promote an SRAD design to a trusted flight critical component, like an altimeter, it must be thoroughly reviewed and tested, both on the ground and on actual flights. Future attempts must keep this in mind.