My ADC read random, noisy values run to run; there was no bug — with the high-impedance buffer enabled and nothing connected, I was watching input bias current charge the amplifier’s input impedance.

I started from someone else’s ADS1255 driver, fixed the original’s mistakes and a few of my own, added some quality-of-life features, and stopped the moment it was good enough to get back to testing the ADC.

Rev 1 of my ADC noise project has one deliberate mistake, a couple of knobs to turn later, and best practice everywhere else — here’s what’s on the board and why.

Building a shell on STM_Shell for the Framework module — what the library gave me, what I had to add for arrow keys and line editing, and the ringbuffer bug that has the project on hold.

Second pass at the Framework bus tool, replacing the bridge IC with an STM32 to push the user interface into firmware and kill the host-software problem.

Setting up usbipd and a few helper scripts to make USB devices reachable from inside WSL, with auto-reconnect when the link drops

First attempt at the Framework bus tool — a clever power topology around a single bridge IC, undone by a software architecture I should have seen coming.

A validation test to compare different planar inductor approximation methods.

Taking a closer look at the Ceedling configuration file to assist with cross-platform development.

Using Gitlab CI/CD tools to automate documentation workflow.