Skip to content

README.md

Latest commit

 

History

History
69 lines (55 loc) · 3.1 KB

README.md

File metadata and controls

69 lines (55 loc) · 3.1 KB
Raw

NaxesAFAP — three-axis FasNAxis example

A hardware example for the multi-axis planner FasNAxis (see extras/doc/n_axes_whitepaper.md). One FastAccelStepper drives each axis independently (X, Y, Z — no gantry); a single FasNAxis planner feeds all axis queues from one polyline so the axes stay time-synchronized.

The path

The path is fixed and deterministic, so a run is comparable to the next:

helix      →  hexagon      →  square      →  return to origin
  • helix — NAXES_HELIX_TURNS full revolutions in the XY plane while Z climbs NAXES_HELIX_Z_PER_TURN per turn, sampled from a quarter-sine table (NaxesAFAP_path.h, float-free).
  • hexagon — six corners of a regular hexagon in the XY plane.
  • square — four corners of an axis-aligned square in the XY plane (the cube footprint on a two-axis build).
  • return to origin — a direct line back to the start position.

The planner runs in Linear mode: each addWaypoint() target is an exact-chord vertex, so the realized curve is the polygon through the vertices. The geometry detector reconstructs that polygon from the Step/Dir pin traces.

Axis count

A compile-time switch (NAXES_HW) drops the third axis on the ATmega168/328/ 328p, which expose only two step channels (OC1A / OC1B). Every other platform wires three axes:

#if defined(__AVR_ATmega328P__) || __AVR_ATmega328__ || __AVR_ATmega168__
#define NAXES_HW 2
#else
#define NAXES_HW 3
#endif

Pin configuration

Following the StepperDemo pattern, the pin table is selected by architecture via StepperPins_NaxesAFAP_<plat>.h (avr, sam, pico, esp32). Each header defines the NaxesAFAP_config_0[] array of stepper_config_s (from StepperConfig.h), with NAXES_HW valid entries plus the STEPPER_CONFIG_END sentinel. Enable is low-active and set manually so it is settled before the planner kicks off (whitepaper §4.5).

Kick-off synchronization

The planner is constructed with the engine (FasNAxis<NAXES_HW, NAXES_HORIZON> NaxesAFAP_planner(FasNAxisConfig{}, engine)) and its kick-off uses the engine's synchronized start: when a committed path is released, every active axis queue is started by one engine operation instead of one addQueueEntry(NULL, true) per axis, so the axes share a single start event. Today each platform implements that operation as a single critical section around the per-stepper starts; platform-specific mechanisms (see extras/todo/engine_synchronized_start.md) replace it as they are written.

  • CI / PlatformIO — pio_dirs/NaxesAFAP/ is a symlink wrapper (like pio_dirs/MoveTimed/): platformio.ini → extras/ci/platformio.ini, FastAccelStepper/src → the library, and src/NaxesAFAP.ino → this sketch. The example is built for every architecture in extras/ci/build_matrix.yaml.
  • SimAVR — extras/tests/simavr_based/test_NaxesAFAP/ builds this example for the ATmega328p (two axes), runs it to completion under run_avr, and the geometry detector validates the reconstructed helix/hexagon/square/origin curve.
  • PC — the planner itself is covered by extras/tests/pc_based/test_26.