The problem
In aerospace repair shops there's a skills gap hiding between two certifications. Plenty of training exists for soldering, and operators get taught to run the automated testers. The diagnostic skill in the middle, actually figuring out why a circuit card fails, is mostly tribal knowledge passed bench to bench. If nobody at your shop has time to teach you, there was no structured place to learn it.
What I built
Circuit Card Academy is a free, self-paced course that takes someone from electronics foundations to independently troubleshooting digital and analog circuit cards in an aerospace repair environment. Eleven modules move from schematics and component identification through multimeter and oscilloscope technique to troubleshooting methodology, automated test equipment, and aerospace workmanship standards. The final module is six worked case files that narrate real diagnostic reasoning: what was checked, why, and where the reasoning could have gone wrong.
Reading isn't practicing, so the course is built around doing: 16 scored fault drills, 8 interactive decision trees, module quizzes with an 80 percent pass bar, a 120-term glossary, and bench calculators. A printable toolkit brings it to the actual bench, with a repair traveler, checklists, pocket cards, and a supervisor sign-off sheet, plus an auto-generated training record for anyone using the course to document competency.
One design decision shaped everything: the course respects its learners. No accounts, no analytics, no data collection. Progress lives on the learner's own device, the whole site works offline, and an employer can self-host it from a single zip file on an air-gapped shop network. Every example board and value in the course is generic and invented, so nothing proprietary or export-controlled ever appears.
What it does now
The course is live and free at circuitcardacademy.com, with a 12-week study plan for self-paced learners and an employer page for shops that want to use it as structured onboarding. It's an independent educational project, not an accredited certification, and it says so plainly on the site.
What went wrong
The honest limitation is scope. I deliberately deferred hands-on soldering and rework practice, regulatory paperwork like 8130 forms, and several failure-mode deep dives, because a web course that pretends to teach hand skills would be worse than one that's honest about needing a bench and a mentor for that part. I also rejected employer dashboards and SCORM integration to protect the zero-backend, no-tracking design, which means shops wanting centralized progress reports have to collect printed records instead. That trade was right, but it was a trade.
What it demonstrates
Instructional design end to end: curriculum structure, plain-language technical writing, interaction design, and the judgment to know what a format can and can't teach.