Ch 14Checkpoint

Checkpoint 14

Checkpoint · Chapter 14 — Calibration and tuning

• Bed and hotend PID tuned in Ch 13 and SAVE_CONFIG'd; both hold within ±0.5 °C at setpoint
• A and B belts both at ~110 Hz over a measured 150 mm span, equal to each other after moving the gantry and returning
• All four Z belts at ~140 Hz over a measured 150 mm span, even with each other — set at the Z idler tensioner bolts, belt clamps untouched
• Closed-chamber soak of 1½–2 h done with panels and door on; QGL run 3–5× hot with shrinking corrections; machinist square clean at both front corners hot; the four Z joint M5×40 bolts tightened hot, idle timeout set back with SET_IDLE_TIMEOUT TIMEOUT=1800 after the last hot check
• QUAD_GANTRY_LEVEL converges in ≤3 retries hot after the lock; PROBE_ACCURACY σ < 0.003 mm with no trend
• rotation_distance verified on the Ch 13 value: 100 mm requested measures 99.5–100.5 mm; 22.6789511 nowhere in printer.cfg
• PRINT_START (Ch 12 Step 12.36, with Step 14.9's purge line) homes, heats the bed, soaks, QGLs, re-homes Z, meshes, heats the nozzle and purges — tested standalone from the console with CHAMBER=0
• Chamber reaches the 50–60 °C band with panels and door closed
• Voron cube (printed in Ch 13) calipered; X, Y within ±0.15 mm and Z within ±0.10 mm of 30.00 mm
• Both cubes (Prusa and Voron) measured into the comparison table
• MEASURE_AXES_NOISE in the 1–100 range; [resonance_tester] probe_points: 175, 175, 20 confirmed live (Ch 12 Step 12.27)
• [input_shaper] saved for X and Y; neither axis below 25 Hz
• [printer] max_accel reduced from the stock 10000 to at or below the calibration's suggestion, with margin
• max_z_velocity: 15 / max_z_accel: 350 restored if Z shaping was attempted
• Pressure advance set per-filament for ASA and recorded
• Extrusion multiplier set in the ASA filament profile and recorded
• First-layer squish re-checked after the extrusion-multiplier change and saved with Z_OFFSET_APPLY_ENDSTOP
• Every fastener the shaper run could have loosened re-checked
• ~/printer_data/config committed (one commit per SAVE_CONFIG) and copied off the Pi
• Tuning log filled in, with initials
• Part H, if fitted: Step 14.25's calculator passed; exactly three jumpers, Fan0, Fan2 and Fan3, all at 24 V; the ASA soak still settles at 50–60 °C with the housing closed; a PLA print held chamber_temp at or under 37 °C, then purged and stopped; config committed

You built a tuned printer with every number in the tuning log. 14 gummy worms.

Common mistakes

• Tensioning belts before Ch 06b squaring. The squaring procedure begins by fully releasing A/B tension; anything set before it is thrown away, and you will have done the work twice (survey §5.2 W1).
• Reading a belt frequency without measuring the span. 110 Hz is only meaningful at 150 mm. Eyeballing the span is the single most common way people end up with belts 30 % out and blame the printer.
• Loosening a Z belt clamp to "adjust" Z tension. The clamps on each gantry corner, on top of the front idler or A/B drive and under the Z joint, hold both belt ends; the adjuster is the idler tensioner bolt at the top of each upright. A released clamp means re-threading the belt.
• Tightening the Z joints cold, or skipping the closed-chamber soak. The value of Step 14.6 is locking the gantry in at full thermal expansion; a cold tighten looks the same on the bench and shows up as first-layer inconsistency.
• Running SHAPER_CALIBRATE before the first successful print, or before final belt tension. Both change the resonances you just measured, so the result is stale before you use it. Same for pressure advance, which also shifts when input shaping is switched on.
• SAVE_CONFIG after the shaper run and stopping there. It writes [input_shaper] but explicitly does not touch max_accel. Leaving the stock max_accel: 10000 means you shaped the ringing and then printed at an acceleration that reintroduces it.
• Fixing an oversize cube with negative XY size compensation or the extrusion multiplier. Negative XY compensation makes every bearing bore and screw hole in the machine oversize, and a multiplier moved to hit a caliper number leaves the top-surface optimum of 14.19–14.20. Voron parts need no compensation beyond a good multiplier tune (Ellis).
• PID-tuning the hotend at 260 °C with max_temp: 270. The relay test overshoots the target and can reach the limit mid-run (verify on bench). Tune at 245, or raise max_temp to 290 first.
• Raising [heater_bed] max_power above 0.6 to speed up warm-up. LDO set it to protect a 355 mm cast plate from warping. Wait the extra five minutes.
• Skipping the heat soak because the bed says 110 °C. The bed reaches temperature in minutes; the frame is there only when PROBE_ACCURACY stops trending. Probing a cold frame gives a mesh and a Z offset that are wrong for every print you then run.
• Soaking on the stock idle timeout. [idle_timeout] timeout: 1800 (LDO config) runs TURN_OFF_HEATERS + M84 30 minutes after the last move; a static soak has no moves, so the heaters go off, the Z motors release and the chamber "stalls". SET_IDLE_TIMEOUT TIMEOUT=7200 before 14.3 and 14.8 (99999 for 14.6, which runs longer), and put it back to 1800 afterwards.

Next

The machine is tuned. Print something for the machine itself — a nozzle brush, a spool holder, the Clicky-Clack spares — and then work Ellis' sections 7 and 8 (cooling/layer times, retraction) as your first real tuning session on ASA.

← PreviousTuning log

Next →Batch B00 — before you start