Test Catalogue

Every claim in this manual is backed by a test that prints the measured value. This chapter lists all tests of the three repositories with their stimulus, their pass criterion and the last result. The Status column says where the result comes from: Scilab (run in Scilab 2027), Python (run in the Python twin, the Scilab run of this version is pending), C (GoogleTest), ngspice.

Naming

Prefix	Meaning	Where
T	test of the model or of the prototype against theory or limits	sfs-stepper-sim/tests/test_T*.sce, twins in tools/
X	cross-check: Scilab equals its Python twin	tests/test_X*.sce + tests/reference/*.csv
B	power stage mapping	tests/test_B01_bridge.sce
S	software in the loop: the C library in Xcos equals the prototype	tests/test_S01_sil.sce
SP	ngspice power stage	sfs-stepper-power-spice/tools/run_tests.py
GoogleTest suites	library unit tests	sfs-stepper-ctrl/tests/unit/

Motor model

Test	Stimulus	Pass criterion	Result	Status
T00	1000 random states; current vector at φ = 0.7	power balance e·i = T·ω within 10⁻⁹ W; zero and restoring torque at p θ = φ	1.1·10⁻¹³ W; −1.0·10⁻⁵ N·m restoring	Scilab
T01	3.2 V on phase A, locked and free rotor	steady current V/R within 10⁻⁶; free rotor stays at θ = 0	2.000000 A, error 2·10⁻⁹	Scilab
T02	voltage step, locked rotor	i(t) = V/R (1 − e^(−t/τ)), t63 = L/R within 0.1 %	t63 = 1.8750 ms	Scilab
X01	rotating voltage 5 Hz el., free rotor	Scilab (lsoda) equals Python (Radau)	3.6·10⁻⁸ A, 8.8·10⁻⁷ rad/s	Scilab

Power stage and current loop

Test	Stimulus	Pass criterion	Result	Status
B01	20 000 random voltage requests, both topologies	duties 0 … 1, direction kept, exact when feasible; largest circle Vbus / Vbus/√2	direction error 2·10⁻¹⁶; 24.0000 V / 16.9706 V	Scilab
T03	microstepping 2 A, 10 Hz el.	A leads B by 90° ± 0.5°, amplitude ± 10 mA, rotor follows	90.000°, 2.0000 A, lag 0.57°	Scilab
T04	current step 0 → 2 A, locked, both topologies	overshoot < 5 %, settling (2 %) < 1 ms, error < 0.1 %	1.12 %, rise 0.35 ms, settling 0.40 ms	Scilab
X02	current step, shared leg	Scilab equals Python	5·10⁻¹³ A	Scilab
T08	controller for the nominal motor; plant R ×0.9 … 1.3, L ×0.8 … 1.2 (current step) and R, L, J ×5, load 0.15 N·m (ramp to 4 rev/s)	step: overshoot < 10 %, settling < 2 ms, error < 1 %; ramp: load angle < 90°	worst settling 1.6 ms (R ×1.3, L ×0.8); worst load angle 46.3°	Python
T09	8 rev/s, both topologies	shared leg limited earlier than full bridge, both synchronous	1.862 A / 1.557 A, load angle 24.5° / 39.3°	Scilab
T10	switching stage: hold 2 A; ±2 A with dead time; 4 rev/s	ideal switches = averaged (< 1 mA); ripple = analytic ± 5 %; dead time error = formula ± 1 %; synchronous, peak < trip	0.01 mA; 23.11 mA p-p (analytic 23.11); 0.07 mV deviation; peak 2.03 A	Python
X04	switching stage, shared leg, 4 rev/s	Scilab equals Python (10⁻⁶ A)	—	pending

Motion

Test	Stimulus	Pass criterion	Result	Status
T05	move 6400 µsteps (2 rev), 5 rev/s, 50 rev/s²	target exact, |Δv| ≤ acc, monotonic, load angle < 45°	499 ms, load angle 15.24°, final 6400.00 µsteps	Scilab
T06	load step 0.25 and 0.5 N·m at 1 rev/s	synchronous below K_t I with angle asin(T/(K_t I)); loss of synchronism detected above	37.21° (theory 37.21°); slip at 151.9 ms	Scilab
T07	reversal +4 → −4 rev/s	synchronous, deceleration limit kept, final speed	load angle 12.25°, −240.0 rpm	Scilab
X03	move 160 µsteps	motion generator + loop equal the Python twin	4.9·10⁻¹³ A	Scilab

Resonance and friction

Test	Stimulus	Pass criterion	Result	Status
T11a	angle step 0.1 rad at 2 A	natural frequency √(p K_t I / J)/2π ± 5 %	248.1 Hz (theory 253.8 Hz)	Python
T11b	speed sweep 0.8 … 1.8 rev/s, detent 0.02 N·m	largest oscillation at 4 p n = f_n ± 10 %, peak > 3× minimum	peak at 1.2 rev/s (theory 1.27), ±13.6°	Python
T11c	5, 8, 11 rev/s, shared leg, 24 V	oscillation < 1° / < 5° below the voltage limit, > 20° beyond	0.0° / 1.3° / 47.6°	Python
T12	unpowered rotor without back-EMF, T_c 0.01 N·m, 50 rad/s	spin-down time (J/B) ln((ω0 + T_c/B)/(ω + T_c/B)) ± 1 %	39.60 ms (theory 39.60 ms)	Python

Software in the loop

Test	Scenario	Pass criterion	Result	Status
S01	current step, locked rotor	currents < 5 mA from the prototype	—	pending re-run
S02	move 640 µsteps	currents < 5 mA, commanded angle identical, rotor < 0.05 µstep	1.05 mA, 0, 0.0075 µstep	Scilab
S03	8 rev/s, both topologies	amplitude as T09 ± 30 mA	1.862 A / 1.557 A	Scilab
S04	run 4 rev/s, switching stage	currents and peak < 5 mA from the prototype	—	pending

Library unit tests (GoogleTest)

Suite	Tests	Covers
SstepMath	29	saturating arithmetic, gains from ratios (brute force over 200 000 cases), sine/cosine accuracy, integrator, duty to ticks
SstepBridge	5	random requests, circular amplitude, bus voltage scaling, very large requests keep the direction, zero request centred
SstepMotion	10	integer square root, two-revolution move (499 ms), short and negative moves, random targets without velocity step, retarget and reversal, run and reverse, hold
SstepApi	24	configuration validation, NULL arguments, state rules, trips, faults, hardware errors, late slow loop, bus voltage corner cases, feed-forward saturation
SstepLoop	8	closed loop with a motor model: current step, move, load step, reversal and stop, voltage limit by topology, bus voltage compensation, voltage drive, other microstep settings

Result: 76 tests pass with the fixed point backend, 75 with float (one test is backend-specific); 100 % line and branch coverage of src/; no finding under ASan and UBSan. Status: C.

ngspice

Test	Pass criterion	Result	Status
SP01	device against the datasheet (Rds ±5 %, capacitances and Qg ±10 %, VSD ≤ 1.2 V)	all within	ngspice
SP02	step model within 15 mV for 1–3 A; driven gates never overlap	12.8 mV; 154 ns gap	ngspice
SP03	shunt sample within 1 %	exact, settling 345 ns	ngspice

Running everything

Repository	Command
Stepper-C	ctest --test-dir "build/[UnitTests]Stepper Motor Control" (and the Float preset)
sfs-stepper-sim	scilab-cli -f tests/run_all.sce; Python: python tools/ref_*.py
sfs-stepper-power-spice	python tools/run_tests.py