Seventeen entries, each verified against a logic analyzer, a power profiler, and an oscilloscope before publication. Filter below — or read in reverse-chronological order, which is how I bench-test them.
RTC memory, GPIO holds, the BOR brown-out trap nobody talks about, and the exact reason your "1µA" board pulls 800µA.
A 24-line WS2812 driver from zero. Includes the logic-analyzer capture of the 800kHz waveform and where the typical example wastes 60% of CPU.
Three lie about their sample rate. One is fine. One is suspiciously good. The fancy one isn't worth it under 24MHz.
Why your multimeter is silently lying below 1mA, and which of these three rigs gives the cleanest log of an MCU wake cycle.
The actual cycle cost of memcpy vs. DMA on G0, F4 and H7 — measured, not estimated. With the Cube config that doesn't hang.
TPS62840 vs. MCP1640 vs. RT8059. Same load curve, same scope, very different ringing. Charts inside.
Why 4.7k is wrong half the time. Includes the rise-time math, three captures at 100k/400k/1M, and the napkin shortcut.
The real attack surface on a Pico W in deployment, OTP fuses, and the failure mode that bricks four boards. Don't skip this.
A printable 24-item checklist drawn from twelve years of catching the same five mistakes. PDF mirror inside.
Side-by-side bench comparisons of the parts and tools that earn shelf space. Same fixture, same load, same logging — the only thing that changes is the device under test.
A working list of references this bench actually reaches for — embedded systems, PCB design, signal integrity, firmware engineering, CAN bus. The books I would hand to a junior engineer asking where to start in each area.
Affiliate links route through Bookshop.org (supports independent bookstores). No book lands here that I would not personally recommend.
Open the library