32-bit-micro-processor (vivasaayi) errors
4 failures reported against 32-bit-micro-processor (vivasaayi), sorted by how they fail and ranked by impact.
#114 by errors trackedElectronics & PCBSimulation & verification
Recurring 32-bit-micro-processor (vivasaayi) errors, by impact
- 1An inventory found that 87 of 88 checked-in KiCad module projects with a top-level PCB have no ERC section in the project file. Only the SHT31 pilot does. The generator copies a template that has explicit ERC config, so somewhere between schematic and PCB generation the settings get dropped. On one board the generator deliberately changes an ERC matrix pair, the saved project does not have it, and KiCad 10.0.8 reports two errors. The author is careful to say the cause is not proven yet.Electronics & PCB · Silent failure · 1 report66
- 2The schematic and the canonical PCB both wire the ECP5's Master SPI chip select to flash IO3, while a different FPGA pin reaches flash CS#. The header says do not fabricate or attempt boot. The corrected schematic and routing exist only as unapproved trials, not merged into the host design, and the fix touches schematic, generator, PCB, pinout, FPGA toolchain and bitstream boot mode at once. A 62 MHz flash clock setting is flagged as unmeasured. Hardware, FPGA config and firmware boot all depend on one pin assignment nobody caught until now.Simulation & verification · Wrong design output · 1 report66
- 3The shared readiness report and fab exporter "could report misleading success, so existing fab-ready counts cannot yet be used as release evidence." The batch checker had hard-coded paths and returned success for failing targets. The fix list is long: fail closed on ERC and DRC, fresh reports per run, build packages in a clean directory so stale artifacts cannot pass, and keep MPN and DNP straight in the BOM export. The fixes are tested only against a simulated KiCad CLI, because kicad-cli is not installed where the work is happening.Electronics & PCB · Wrong verdict · 1 report66
- 4The TPS25751 USB-PD controller's schematic models every power terminal separately, while its PCB footprint groups them into shared lands. KiCad reports seven parity conflicts that remain, and the owner wants each one resolved or dispositioned without concealing electrical differences, plus a review of the VCONN supply that is currently left unpowered. The schematic and the footprint describe the same chip two different ways, and somebody has to decide which one is true.Electronics & PCB · Wrong design output · 1 report66
How it fails
Sourced from public community reports. Costs and delays on individual entries are modeled estimates, not measurements.