unisat

UniSat — Characterisation data templates

This directory holds the measured characteristics of the firmware running on the real STM32F446RE target hardware. The numbers the design documents quote (168 MHz clock, ≤ 60 % CPU load at 10 Hz telemetry, 5 krad TID tolerance of the STM32F446, …) are calculated or vendor-claimed; the files here hold measured values from a specific OBC unit, identified by serial number.

For Phase 5 / TRL-5 the characterisation data itself is not yet populated — this directory establishes the format so that measurements can be filed systematically when the HIL bench (see docs/testing/hil_test_plan.md) is brought online. Each template file describes what measurement goes where, what units, and which tool produced it.

Templates

File	Measurement	Source	Tool
wcet.md	Worst-case execution time per FreeRTOS task	SWV + DWT cycle counter	ST-Link / openocd + manual analysis
stack_usage.md	FreeRTOS task stack high-water mark	uxTaskGetStackHighWaterMark() over 48 h soak	firmware TLM downlink
heap_usage.md	heap_4 free bytes min/max/fragmentation	xPortGetFreeHeapSize(), dump post-soak	firmware TLM downlink
flash_ram_footprint.md	.text/.data/.bss section sizes, %-of-capacity	arm-none-eabi-size	Phase 1 build-time gate
power_profile.md	mA draw per mode (idle / nominal / imaging / safe / TX)	bench USB ammeter + scope	manual
boot_time.md	µs from reset to first FreeRTOS task spin	SWV trace with timestamp	manual
i2c_bus_timing.md	SCL period, ACK latency per sensor	logic analyser	Saleae / OWON
rf_link_budget.md	Measured BER vs SNR at 437 MHz	RTL-SDR loopback	direwolf

Policy

• One file per metric. Each file states: measurement units, setup diagram, repetition count, results table, and the date + OBC serial.
• Never delete. When a firmware change invalidates an earlier measurement, prepend a new section with the new value and annotate the previous section as *superseded by §N of commit *. Rolling history is the audit trail reviewers look for when validating a TRL-5 claim.
• Keep raw data separately. Reference large traces / CSVs / scope captures by path and SHA-256, don’t commit the blob. Branch-policy target ≤ 50 MB.

Gap versus a genuine TRL-5 claim

A honest TRL-5 characterisation package must cover the four environments listed below. The files in this directory are the software-side bookkeeping; the hardware-side proof (vibration, thermal-vacuum, radiation) lives outside the repository and is referenced by external report number.

Environment	Measurement needed	Where reported
Room-temperature bench	All .md files in this directory	this repository
Thermal-vacuum	Functional telemetry + beacon cadence across −40..+80 °C sweep	external TVAC report
Vibration (launch profile)	Pre/post bench-run telemetry equivalence	external vibration report
Radiation (TID ≥ 10 krad)	Post-irradiation bench-run delta	external cyclotron report

The external-report fields are placeholders for operator-supplied PDFs. The repository does not claim TRL 5 without them — see docs/project/GAPS_AND_ROADMAP.md §”Out of scope” for the explicit disclaimer.

This site is open source. Improve this page.