Logan Brown
Attending a project based school has allowed me the freedom to be continually curios, creative, and and solve real-world problems.
​
In this project, the class was split into teams of 2-3 people. My 2 person team was given the randomly selected topic of Digital Photography and instructions to research the topic, create a poster, and present to the class. In addition, we asked ourselves, "Is it possible to design, fabricate, and test a functional digital camera prototype in the 3 weeks given to complete the assignment?" After receiving permission from the teacher to alter the deliverable goals, this is the result.
3 Weeks
Designing, fabricating, and testing a functional digital camera prototype in the 3 weeks was a monumental task. To complete this we had to work on many different aspects of the project in parallel, with tight communication and planning. We began by discussing possible solutions and minimum deliverable requirements.
​
Requirements:
-
10 x 10 pixel resolution image sensor
-
Control board to receive and process signals
-
Software to decode and display images
-
A simple focal system for the image sensor
​
Image Sensor:
To save time and reduce cost, image sensor board design, and component sourcing and purchasing was done in parallel. We selected photo resistive diodes sensitive to infra-red (IR) light used in television remote controls. While the printed circuit board (PCB) for the image sensor was being designed we began searching vendors for all the components needed with the best possible price and availably to arrive with plenty of time to complete the build. We used KiCAD for design and JLCPCB.com to manufacture the PCB. Components were sourced from various vendors based on price, availability, and shipping speed.
​
Control Board (Image processor):
We began designing and programming the image processor while we waited for the image sensor PCBs and components to arrive. We chose to use a blue-pill STM32 microcontroller due to it's very low cost, good processing power, and availability. Arduino boards were considered as well because of their ease of use but were excluded due to cost.
​
Display Software:
Display software and Control Board software were done in parallel. This required a lot of communication and collaboration between team members to ensure data was formatted and provided from the control board to the PC based display software in the correct format and order.
​
Assembly Problems:
Once the Image sensor PCB and components arrived we began assembly. This did not go smoothly. Due to tight time constraints in the design phase, we discovered a critical flaw in the Image Sensor PCB design. The photo resistive diodes were not compatible with the STM32 control board. With 1 week left in our timeline we did not have enough time to update and resend the image sensor PCB for manufacture. Rather than give up, we brainstormed a solution and, constructed a custom wiring harness and proto board with additional components we had on hand to correct the problem. While this made the design less elegant, it worked.
​
Additionally, after soldering the IR diodes on to the PCB we discovered 1 row of diodes placed in the wrong polarity. Rather than redo all 200 solders on a fresh board which would be time consuming. We carefully de-soldered the bad row and flipped the polarity. Luckily we had ordered enough spare diodes in case we encountered a problem later on.
​
Display Software Testing:
After getting the sensor and control board operational we began testing and iterating the display software. This went relatively smoothly and was shared between team members. This kept us both busy and productive at all times right up to the presentation. Two things we discovered during testing were that we could capture video as well as still images which exceeded our minimum requirements. We also discovered that the image quality was highly dependent on the amount of available IR light. This prompted us to quickly build an IR LED matrix to improve the quality of images and video displayed for our presentation.
​
Coming into Focus:
After testing our original focal system based on an insect eye, we found that it had an extremely limited focal range. Rather than accept this we quickly pivoted and redesigned the entire focal system around a standard lens we scavenged from an old science kit, and added an adjustable focal length mechanism with a custom laser-cut and 3d printed housing.
​
Finished Prototype:
We managed to complete a functional digital camera prototype in 3 weeks. Along the way we learned many new skills and collaborated closely and effectively. I'm thankful to have teachers, team members, and an environment that has allowed me to expand what's possible and go beyond what is expected.