RailCore® is a family of Core-XY based Reprap 3D printers designed by J. Steve White, Tony Akens, and Ben Withem.
This walks through a Klipper printer.cfg for a RailCore, section by section. The
sample configs put it all together:
Every value marked CHECK: in the samples must be verified on your printer. They come from the
kit RRF configs and the build guides, and RailCores differ: self-sourced parts, different probe
mounts, different hotends. Never copy heater limits, PID values or input shaper numbers from
another machine.
New to Klipper? Read the Klipper landing page first.
Do these in order, one at a time. Klipper’s Config checks page has the details for each.
[mcu] serial: (ls /dev/serial/by-id/* on the host), then FIRMWARE_RESTART.QUERY_ENDSTOPS with each switch released and pressed (X, Y, probe).STEPPER_BUZZ STEPPER=stepper_x (and stepper_y, stepper_z, stepper_z1, stepper_z2,
extruder): confirm each motor moves and which way.PID_CALIBRATE HEATER=extruder TARGET=240
and PID_CALIBRATE HEATER=heater_bed TARGET=100, then SAVE_CONFIG.Z_TILT_ADJUST, PROBE_CALIBRATE, a mesh, input shaping, then your first print.The build guide’s S) First Moves (Mini: S) does the same checks for RRF; the expected results are the same.
[printer]
kinematics: corexy
max_velocity: 400
max_accel: 3000 # raise after input shaping
max_z_velocity: 15
max_z_accel: 100
The RailCore is a standard CoreXY. On the RailCore II (build guide J and Q):
stepper_x.stepper_y.position_endstop: 0.On the RailCore Mini (guide Q) X homes to
max (endstop at the rear right) and Y to min, so stepper_x uses
homing_positive_dir: true and position_endstop: 150. The kit config.g maps X to drive 6 and Y
to drive 5, but the guide’s wiring step labels them the other way round: check with the motion test.
Steps: RRF’s steps/mm becomes rotation_distance (mm per motor turn).
Klipper’s rotation distance guide explains it.
| kit RRF | Klipper | |
|---|---|---|
| X/Y, 16T pulley, 0.9° motors | M92 X200 Y200 |
rotation_distance: 32, full_steps_per_rotation: 400 |
| Z, TR8x4 leadscrew, 0.9° motors | M92 Z1600 |
rotation_distance: 4, full_steps_per_rotation: 400 |
| Z, TR8x2 leadscrew, 0.9° motors | M92 Z3200 |
rotation_distance: 2, full_steps_per_rotation: 400 |
| Extruder, RailCore II kit BMG, 0.9° motor | M92 E837 |
rotation_distance: 7.646, full_steps_per_rotation: 400 (calibrate) |
| Extruder, Mini kit LGX Lite, 1.8° motor | M92 E562 |
rotation_distance: 5.694, full_steps_per_rotation: 200 (calibrate) |
The general rule: rotation_distance = full_steps_per_rotation × microsteps ÷ steps_per_mm.
Motor current: RRF’s M906 is in mA, Klipper’s run_current is in amps RMS. Don’t copy the
number across: set it from your motor’s rated current (about 70% of rating is a sane start) and
check the motors stay below about 70 °C. The kit LDO X/Y motors (42STH47-1684MAC) are rated 1.68 A.
Directions: RRF’s M569 S0/S1 does not map one-to-one onto Klipper’s dir_pin. If a motor runs
the wrong way, add or remove ! in front of its dir_pin. If X moves when you asked for Y, or the
axes move diagonally, the X/Y motor plugs are swapped or one direction is wrong.
z_tiltThe RailCore II and the Mini both lift the bed with three independent leadscrews, each on its own
driver. Klipper’s [z_tilt] probes the
bed and moves each motor to level it: the same job as RRF’s M671 plus G32/bed.g.
Define the motors as [stepper_z], [stepper_z1] and [stepper_z2]. Only stepper_z carries the
endstop and travel settings. The order matters: the first line of z_positions belongs to
stepper_z, the second to stepper_z1, the third to stepper_z2.
RailCore II kit (build guide H and Q; two motors on the left, one on the right):
| Klipper | Duex 5 drive | position | kit RRF M671 |
|---|---|---|---|
stepper_z |
5 | front left | X-10 Y22.5 |
stepper_z1 |
6 | rear left | X-10 Y277.5 |
stepper_z2 |
7 | right | X333 Y150 |
[z_tilt]
z_positions:
-10, 22.5 # stepper_z front left CHECK: measure yours
-10, 277.5 # stepper_z1 rear left
333, 150 # stepper_z2 right
points: # NOZZLE positions; the probe lands at point + probe offset
20, 0
20, 230
270, 108
speed: 150
horizontal_move_z: 5
retries: 5
retry_tolerance: 0.0075
RailCore Mini kit (one motor on the left, two on the right):
| Klipper | Duet 3 Mini drive | position | kit RRF M671 |
|---|---|---|---|
stepper_z |
4 | left | X-5 Y90 |
stepper_z1 |
3 | rear right | X165 Y155 |
stepper_z2 |
2 | front right | X165 Y22.5 |
Notes:
z_positions are the leadscrew (pivot) positions in nozzle coordinates, so they sit outside
the bed. The kit config says to measure “to the bolt for the yoke of each leadscrew”. The kit
numbers are a good start; one community 300ZL measured -11,20 / -11,274 / 347,148. Small errors
only cost an extra retry.points are nozzle positions. Klipper moves the nozzle there and the probe lands at the point
plus x_offset/y_offset. Put one point close to each leadscrew, and check every point puts the
probe on the bed: the PROBE_LEADSCREWS macro in the sample visits each one.Z_TILT_ADJUST only has small corrections to make.Z_TILT_ADJUST makes the tilt worse on every retry. Stop, fix its
dir_pin, level by hand, try again.G32 in the sample macros runs G28, Z_TILT_ADJUST, G28 Z, the same name as on RRF.The kits use a BLTouch. RailCore II kit wiring: signal to the Duet Z_Probe IN, servo wire to Duex 5 PWM1. Mini kit: both on io3.
[bltouch]
sensor_pin: ^PC1 # Duet 2 Z_Probe IN
control_pin: !PC3 # Duex 5 PWM1 CHECK
x_offset: 2 # kit RRF G31 X2 Y42 CHECK: depends on your mount
y_offset: 42
samples: 2
samples_result: median
samples_tolerance: 0.02
samples_tolerance_retries: 3
G31 X Y and Klipper’s x_offset/y_offset use the same sign: where the probe is,
relative to the nozzle. Measure yours: the kit mount, Mandala Rose Works mounts and printed
mounts all differ.z_offset with PROBE_CALIBRATE (paper test), then SAVE_CONFIG. Do not guess it.PROBE_ACCURACY should give a range of a few hundredths of a millimetre. If not, fix the mount
before tuning anything else.[homing_heaters] turns the hotend heater off during probing, like RRF’s M558 B1, so heater
noise doesn’t trigger the BLTouch.[probe] or the
probe’s own Klipper module. See Klipper’s Probe calibrate guide.X and Y home to their switches. Z homes with the probe, so Klipper needs to know where to put the head first:
[stepper_z]
endstop_pin: probe:z_virtual_endstop
position_min: -2
[safe_z_home]
home_xy_position: 148, 108 # nozzle position that puts the probe at the bed centre
z_hop: 5
home_xy_position is a nozzle position: subtract the probe offset from the point you want to
probe. On the Mini the kit homez.g uses nozzle 73,33.
The RailCore’s MIC6 plate is flat, and z_tilt takes out the tilt, so a mesh mostly corrects small
leadscrew and frame errors. Use Klipper’s adaptive mesh: it only probes under the part.
[bed_mesh]
mesh_min: 12, 42 # PROBE coordinates (not nozzle)
mesh_max: 280, 280
probe_count: 7, 7
algorithm: bicubic
fade_start: 1
fade_end: 10
adaptive_margin: 5
[exclude_object]
z_tilt points, mesh_min/mesh_max are probe coordinates. The kit RRF mesh was
M557 X10:280 Y50:280.BED_MESH_CALIBRATE ADAPTIVE=1 needs [exclude_object] and “Label objects” turned on in your
slicer. Without it, Klipper probes the full grid.Z_TILT_ADJUST before meshing. See Klipper’s
Bed Mesh guide.Klipper measures the printer’s resonances and picks a shaper. This is usually the biggest quality
gain over a stock setup, and it lets you raise max_accel.
[mcu rpi], [adxl345] and [resonance_tester] sections in the sample and set
probe_points: 150, 150, 20 (Mini: 75, 75, 20).ACCELEROMETER_QUERY, then SHAPER_CALIBRATE, then SAVE_CONFIG.The X and Y frequencies differ between RailCores (belt tension, carriage mass, hotend, panels), so measure your own. Re-measure after changing belts or the toolhead.
B4725 C7.06e-8) is
ATC Semitec 104GT-2; a Slice Engineering thermistor is SliceEngineering 450; the E3D Revo
(Mini kit) is ATC Semitec 104NT-4-R025H42G. The kit bed thermistor (B3950) is Generic 3950.
The Mini’s bed (B4240) needs a custom [thermistor] section; the sample has one.max_temp: the kit RRF limit for the hotend is 285 °C (M143 S285). Never set it above what
your hotend, thermistor and heater are rated for.PID_CALIBRATE for both heaters before
printing.[verify_heater] at Klipper’s defaults. It shuts the printer down if a heater stops
heating as expected, the job RRF’s heater fault detection does. Fix the cause, don’t loosen it.[temperature_sensor] for monitoring.[heater_fan] above 50 °C, as in the kit config.The shared macros file has:
G32: G28, Z_TILT_ADJUST, G28 Z (same name as on RRF).PRINT_START BED=60 HOTEND=210: heat the bed, home, Z_TILT_ADJUST at temperature, adaptive
mesh, heat the nozzle, purge line.PRINT_END: retract, lift, park at the rear left, heaters and motors off.PROBE_LEADSCREWS: visits each z_tilt point so you can check the probe lands on the bed.Slicer start G-code (PrusaSlicer, SuperSlicer, OrcaSlicer):
PRINT_START BED=[first_layer_bed_temperature] HOTEND=[first_layer_temperature]
Cura: PRINT_START BED={material_bed_temperature_layer_0} HOTEND={material_print_temperature_layer_0}.
End G-code: PRINT_END. Remove the slicer’s own heat-up and homing lines so they don’t run twice.
PAUSE, RESUME and CANCEL_PRINT come from Mainsail’s mainsail.cfg or Fluidd’s fluidd.cfg;
include whichever your installer created.
rotation_distance: extrude 100 mm and measure (Rotation distance guide).max_accel toward what SHAPER_CALIBRATE recommends.Questions: the #help channel on the RailCore Discord. If you find a value in the samples that is wrong for a stock kit, open an issue or PR on railcore/configs.