Slicer setup — PrusaSlicer profile and Voron overrides

One-time setup for every print batch (B00–B10). Read this chapter before B00; batch chapters link back here and list only their own deviations (brim, orientation, accent colour). B11, the PETG V0 bay ducting, uses its own bundle, slicer/bay-ducts-petg.ini, and the textured sheet: see its page.

Set the profile once, carefully. Twenty two plates inherit these numbers and none of them ask twice.

Slicer: PrusaSlicer 2.9.6 (current stable, released 2026-06-25). PrusaSlicer 3.0.0-alpha12 (published 2026-09-21) is an early preview. (releases)

Version posture. First build: 2.9.6 for every batch, B11 included (decision 2026-09-26). The 3.0 preview is not used for this build; PrusaSlicer 3.0 preview at the end is reference only, for when it goes stable. Changing slicer version is a toolchain change: re-run Gate A — and Gate B once the kit is here — before the next plate.

Base profiles (read out of a PrusaResearch.ini with inherits resolved). Two copies of that file exist and they are not the same file:

• /Applications/PrusaSlicer.app/Contents/Resources/profiles/PrusaResearch.ini — the copy shipped inside the 2.9.6 app bundle, config_version = 2.4.14. This is what slicer/resolve_preset.py reads, so it is the base the committed .ini and the plate 3MFs were derived from.
• ~/Library/Application Support/PrusaSlicer/vendor/PrusaResearch.ini — the vendor bundle the GUI actually resolves presets against, auto-updated in place. It is at config_version = 2.5.9 today (2.5.9, 2026-09-08, only added INDX/XL filament profiles; nothing for the CORE One HF0.4 or Prusament ASA changed). Upstream is at 2.5.12 (2026-10-15: INDX tool-change script, Generic PVA, INDX firmware notice; 2.5.11 on 2026-10-02 added HT hotend support for the CORE One/L; 2.5.10 on 2026-09-17 added Prusament PLA Lightweight profiles and 0.25 mm XL/INDX print tweaks). Resolved through inherits, the CORE One HF0.4 printer, its start G-code, the 0.20mm STRUCTURAL print, Prusament ASA and Prusament PETG V0 carry the same values in both, and 2.5.12 leaves them and Prusament PETG unchanged; 2.5.11 only rewrites Prusament ASA's compatibility conditions to keep it off HT-hotend profiles, and it still matches the HF0.4 printer. The update is safe to take.

They differ on one key that reaches these plates: 2.5.8+'s CORE One start_gcode bumps the firmware check from M115 U6.5.3+12780 to U6.8.1+16182. Nothing mechanical differs; a re-derive should be resolved against the live bundle. Our plates carry neither version verbatim — every one of them ships the cold-probe start G-code instead (below), and that is what makes the Plater's printer box read "(modified)".

The machine has the high-flow 0.4 nozzle, so the HF variants are the base wherever one exists:

• Printer: Prusa CORE One HF0.4 nozzle (printer model Prusa CORE One & CORE One+) — bed 250×220, max height 270, retract 0.7 mm @ 45 mm/s, z-hop 0.2 mm, wipe off. The non-HF Prusa CORE One 0.4 nozzle preset inherits from this one and only sets nozzle_high_flow = 0; every mechanical value is identical.
• Filament: Prusament ASA @COREONE HF0.4 — nozzle 265 °C, bed 110 °C, chamber 55 °C (minimum 40 °C), fan 20–25 %, first 4 layers fan off, density 1.07 g/cm³, max volumetric flow 26 mm³/s. (The non-HF Prusament ASA @COREONE is 260 °C / 15 mm³/s.)
• Print: 0.20mm STRUCTURAL @COREONE 0.4 — not SPEED. STRUCTURAL is already the quality-biased profile (perimeters 70 mm/s vs SPEED's 170; external 50 vs 170; infill 120 vs 200).

Caution

There is no 0.20mm STRUCTURAL @COREONE HF0.4. At 0.20 mm the bundle ships STRUCTURAL only in the non-HF variant; the HF entries at this layer height are SPEED and BALANCED. That is not a problem: 0.20mm STRUCTURAL @COREONE 0.4's compatibility rule is printer_model=~/(COREONE|…)/ and nozzle_diameter[0]==0.4 with no nozzle_high_flow clause, so PrusaSlicer offers it on the HF printer. Do not "upgrade" to 0.20mm BALANCED @COREONE HF0.4 — it inherits from the STRUCTURAL preset and then raises perimeter 70→150, external 50→200 and small-perimeter 50→170 mm/s, and drops bottom solid layers to 3.

High flow raises the ceiling, not the speeds. Every speed below stays exactly as the table says. At those speeds the peak volumetric flow is 110 mm/s solid infill × 0.4 × 0.2 = 8.8 mm³/s, well under the 15 mm³/s the non-HF profile allows, let alone 26 — so the HF base changes the melt temperature and the start G-code and nothing else about these plates.

Committed projects — open the plate, don't rebuild it

You do not re-enter any of this by hand. Every plate is committed as a PrusaSlicer 2.9.6 project with the arrangement, the per-object brims and the full configuration already inside it:

What	Where
One project per plate, 22 of them	slicer/plates/B01-P1.3mf …
Config bundle for --load (black / accent)	slicer/voron-coreone-asa.ini, slicer/voron-accent-blue.ini
The cold-probe start G-code, vendored	slicer/coreone-cold-start.gcode (slicer/sync_start_gcode.py keeps the inis and the 22 projects in step)
Every override below → ini key → value → the line it came from	slicer/OVERRIDES.md
SHA256 + pinned commit of every STL	slicer/stl/MANIFEST.sha256
Sliced time and grams per plate, vs the old model	slicer/estimates.csv
The pre-B00 hot first-layer check (five squares, not a plate of the run, outside every total)	slicer/checks/hot-first-layer.3mf and hot-first-layer-heat-off.3mf, the same plate on slicer/coreone-heat-off-start.gcode (slicer/hot_check.py rebuilds and checks both)

Open slicer/plates/<plate>.3mf in PrusaSlicer (File → Open Project) and slice. The project carries its own print/filament/printer configuration, so it does not matter which presets you had selected. After a plate is re-saved from the GUI, refresh its numbers with python3 slicer/build_plates.py --from-3mf <plate id> [<plate id> ...]: it re-slices those committed 3MFs as they are, replaces their rows in slicer/estimates.csv (and their run's total) and redraws their diagrams. --from-3mf with no ids does the same for all 27 plates, which is safe but slow. Never drop --from-3mf: without it the script re-packs the plates from slicer/plates.py and rewrites their 3MFs, all 27 when no ids are given, and every hand arrangement in them is lost. python3 slicer/fetch_stls.py re-downloads the pinned STLs if slicer/stl/ is empty.

Caution

The plates carry the cold-probe start G-code — leave it alone. All 22 projects and both .ini bundles ship the same block as the GUI preset Prusa CORE One HF0.4 nozzle - coldstart: the nozzle stays at 0 °C through homing and the whole mesh, and the G29 P9 nozzle wipe is removed because a cold tip cannot be wiped. Clean the tip by hand while it is hot, after each print. Opening a project therefore shows the Printer box as Prusa CORE One HF0.4 nozzle (modified) — that modification is the cold start.

One-time PrusaSlicer setup — before the first project

1. Configuration → Configuration Wizard → Prusa FFF → Prusa CORE One & CORE One+ → tick the 0.4 HF nozzle (tick it even if plain 0.4 is already ticked) → Filaments: Prusament ASA → Finish. A fresh install without the HF variant makes PrusaSlicer ask what to do about the project's missing printer profile when you open a plate, and the wrong answer maps the plate onto whatever printer is active. The estimate check in every Load step would catch that, but only after a wasted quarter hour.
2. File → Open Project → slicer/plates/B00-P1.3mf. If a dialog offers to install a printer profile, accept.
3. Read the three preset boxes on the Plater tab. They will not show the bare names above:

Box	What it shows	Why
Printer	Prusa CORE One HF0.4 nozzle — always with (modified)	the project replaces start_gcode with the cold-probe block and adds one printer key, thumbnails (slicer/build_config.py, a 640×480 PNG so a GUI G-code export carries a doc-sized preview). Both are expected; the "(modified)" is the cold start and must stay
Print	0.20mm STRUCTURAL @COREONE 0.4 (modified)	the project's overrides sit on the system preset
Filament	Prusament ASA @COREONE HF0.4 - Voron black (- Voron blue on the three B02 plates)	the project's renamed filament preset, shrinkage zeroed

"(modified)" on the printer box is not a fault and is not the signal to look for — it is the cold-probe start G-code, and a printer box without it means the project did not load. The other two signals that it did not load are a print box without "(modified)" and a filament box reading the bare Prusament ASA @COREONE HF0.4 — any of those, or an estimate that does not match the Load step's numbers, means do not slice; reopen the project. A printer box naming a different printer entirely (not just "(modified)") is the wizard problem in step 1.

You do not need to select the GUI preset Prusa CORE One HF0.4 nozzle - coldstart before opening a plate: the preset and the project hold the identical block, and the project's own copy wins. Selecting it changes nothing, and never "reset to system value" on the printer box — that throws the cold start away.

The times and weights in every batch chapter are PrusaSlicer 2.9.6 estimates from these projects, not a throughput model. Treat the first plate as the calibration of the printer, not of the estimate: if B00-P1 comes off within a few minutes of 4.0 h, the whole table is trustworthy.

Overrides

"Source" = where the number comes from. A source of judgment means it is a call made for this build, not a cited spec — treat it as adjustable if the calibration cube (below) says otherwise.

Print settings

Setting	STRUCTURAL default	Set to	Why	Source
Layer height / first layer	0.20 / 0.20	keep	Voron spec	Manual p.4
Perimeters	2	4	Voron spec; wall count carries load in these parts	Manual p.4
Top solid layers	5	keep 5	Voron spec	Manual p.4
Bottom solid layers	4	5	Voron spec says 5 top and bottom	Manual p.4
Fill density	15 %	40 %	Voron spec	Manual p.4
Fill pattern	Grid	keep Grid	Voron allows grid/gyroid/honeycomb/triangle/cubic	Manual p.4
Extrusion width (default, perimeter, external, infill, solid)	0.45	0.40	Voron: "Extrusion width — Recommended: Forced 0.4 mm". Ribs, bosses and bearing seats were drawn around a 0.4 bead.	Manual p.4
First layer extrusion width	0.50	keep 0.50	wider first bead = better ASA adhesion	judgment
Top infill extrusion width	0.42	0.40	consistency	judgment
Perimeter generator	Arachne	keep Arachne	Voron's "forced 0.4" predates Arachne. Arachne varies bead width on purpose and handles Voron's thin ribs better than Classic. Setting nominal width to 0.4 gives Arachne the right target.	judgment
Supports	on (auto off)	None	every Voron/LDO/Nevermore STL is pre-oriented with built-in break-away supports where needed	Manual; Nevermore README; Klicky STL README
Seam position	Aligned	Rear	keeps the seam off the visible outward faces of skirts and the toolhead	build spec
Skirt loops	0	1 loop, 3 mm gap, min length 4 mm	primes after the long ASA purge; lets you abort in the first 60 s if the first layer is wrong	judgment
Brim	off	off globally; 5 mm on the tall/narrow parts, 3 mm on the 150–182 mm skirts — already applied per object in the committed 3MFs, see §Orientation & brim below	ASA corner lift on long flat parts and tippy tall parts	Prusa warping KB; judgment
XY size compensation	0	keep 0	Voron: "The parts have been designed with ABS/ASA shrinkage in mind… shrinkage should be set to 100 %. Compensating for this is likely to make bearing fits and screw holes too large."	docs.vorondesign.com/materials.html
Elephant-foot compensation	0.20	keep 0.20 for now, verify on the cube	shrinks only the first layer; several Voron parts have a bearing bore starting at the bed — measure before changing	judgment
External perimeter speed	50	35	surface finish and corner accuracy on the visible skirts; speed is not a goal here	judgment
Perimeter speed	70	55	more time at temperature per bead = better interlayer bond at 4 walls	judgment
Infill speed	120	100	interior quality and less pressure variation into the perimeters	judgment
Solid infill speed	140	110	flatter top/bottom faces on the skirts	judgment
Top solid infill	80	keep 80	already conservative	—
First layer speed	45	25	ASA on smooth PEI is the single biggest failure mode; slow the first layer (and see §Print sheet)	Prusa ASA KB; judgment
External perimeters first	off	keep off	inside-out gives better dimensional accuracy on holes; only flip it if you see external-perimeter bulging	judgment
Dynamic overhang speeds	on (15/25/45/90 %)	keep on	Prusa tuned these for the Core One's part fan	—
Ironing	off	keep off	—	—

Filament settings (Prusament ASA @COREONE HF0.4)

Setting	Profile default	Set to	Why	Source
Shrinkage compensation XY	0.22 %	0 %	Same rule as XY size compensation — PrusaSlicer 2.9 scales ASA parts up 0.22 %; that is +0.15 mm on a 66.7 mm Z-drive body and +0.035 mm on every 16 mm bearing bore, compounding with any flow error. This is the single most important override in this document.	docs.vorondesign.com/materials.html + profile value
Shrinkage compensation Z	0.22 %	0 %	same reason	as above
Nozzle / first-layer nozzle	265 / 265 °C	keep	Prusa's own Prusament ASA values for this machine and this nozzle — the HF0.4 profile runs 5 °C hotter than the non-HF one	profile (Prusament ASA @COREONE HF0.4)
Bed / first-layer bed	110 / 110 °C	keep	Prusa ASA guidance is ≥100 °C bed	profile; Prusa ASA KB
Chamber temperature	55 °C	keep 55	matches the 55–60 °C a real Voron sees; the Voron parts are designed for it	profile; docs.vorondesign.com/materials.html
Chamber minimal temperature	40 °C	keep 40	printer will not start until the chamber reaches 40 °C — this is the preheat gate	profile
Min / max fan	20 / 25 %	keep for B00–B02, then decide	Prusa tuned this for the Core One chamber. If perimeter separation shows on the Z-drive bodies, drop to 0 / 15 % and keep bridge fan at 25 %.	profile; Ellis — Perimeter Separation
Max volumetric speed	26 mm³/s	keep	not a limit at these speeds — the plates peak at 8.8 mm³/s	profile (HF0.4)
Retraction / z-hop	0.7 mm / 0.2 mm	keep	Nextruder-specific; do not touch	printer profile

Drying

ASA is mildly hygroscopic. A fresh, sealed Prusament spool used within ~2 weeks needs nothing. Dry at 80 °C for 4 h (Prusa KB) if a spool has been open longer, or the moment you see stringing, matte striping, or popping. The Sunlu SP2 on the bench tops out at 70 °C, so there it is 70 °C for 6 h, with the port plugs out. Keep the active spool in one USS Drybox during the run. Voron parts are structural — this is the material where wet filament actually costs strength. (Prusa ASA KB.)

Print sheet

Decided, and written on the sheet's edge tape; every batch's pre-print step then says "sheet per 00-slicer-setup" and nothing more.

• Which sheet, by material. Three sheets, one job each; the sheet follows the material, never the part:

Material	Sheet	Why
ASA (every Voron plate, the add-on prints)	smooth PEI, with the glue film below	Prusa's ASA sheet; on satin ASA grips less and often needs a brim, and textured holds it poorly
PLA (bench prints: hardware trays, gauges, jigs)	satin	Prusa: satin gives "optimal adhesion, especially when printing PLA and PETG" (satin sheet KB)
PETG, PETG V0 (B11)	textured	PETG bonds too hard to smooth PEI and can tear it; textured releases it cleanly

Any other material: check Prusa's material table before the first print, and add a row here. On the printer, select the matching sheet profile before the first layer, because Live Adjust Z is kept per sheet (verify on bench). - Glue stick under ASA on the smooth sheet. ASA bonds to PEI hard enough that a big flat part (the 66 mm Z-drive bodies, the 182 mm rear skirt) can pull the coating off a smooth sheet; the glue is a separation layer, not an adhesion aid. Thin, even film over the printed area, re-applied every 2–3 plates. - Cleaning: wash the old glue off with warm water and dish soap, then IPA on the smooth sheet. Never acetone or ASA juice on a powder-coated sheet — it strips the coating.

Tip

Prusa's own ASA article is the source for the sheet, the glue stick and the acetone warning — "use the smooth or powder-coated PEI sheet with a glue stick", "ASA sticks very well to our print sheets". help.prusa3d.com/article/asa_1809

Calibration sequence — run before B00, and again after any toolchain change

The dimensional gate is in two parts. Gate A needs only the cube. Gate B tests the fits, and the parts it would fit against — the 625-2RS bearing and the MGN12 rail — come in the Voron kit. Nothing is bought to bring them forward: what runs early is the dimensional half of Gate B, a caliper on the printed bores and seven real inserts out of the KADRICK kit already on the bench. The bearing press and the rail slide are signed off on kit day, and the rail row gates nothing but a 20-minute reprint of the ASA MGN12_rail_guide (3 g). Each gate releases its own set of batches (table below).

Printer status, 2026-09-23. Bed flatness is closed — flat to ±0.1 mm on printed Z-stop correction caps, verified at 60 °C. Belt tuning was never formally closed; the Gen 2 belt upgrade that would redo it is now deferred to the INDX rebuild this winter (Prusa documents the combined install), so the whole 157.0 h run stays on the current Gen 1 GT2 belts. Run this sequence after two pre-B00 checks instead — see B00 § Before B00 checks: a belt pluck check (belt tuning is still open) and a hot first-layer check at ASA bed temperature (the bed has only been verified at 60 °C), run on its committed project right after Step B00.0 installs the presets. The Gen 2 expansion joints and nozzle wiper go in first (ruling 2026-10-01); only the belts wait for INDX.

1. Advanced Filtration Kit — already fitted (2026-09-12), so nothing to install here. ~157 h of ASA is about to run in an enclosure: confirm on the first ASA plate that the blower runs and the bypass flaps seal. One spare cartridge is still to buy (~600 print-h each, 1–2 wk Prusa lead time).
2. Firmware ≥ 6.8.1 on the Core One+, required either way. It runs 7.0.0 (2026-10-05). 6.8.1 carries the Settings → Hardware → Edition selector and the individual Gen 2 toggles the rebuild will use; that 7.0.0 keeps both is (verify on bench).
3. First layer: there is nothing to run. The Core One+ has no first-layer calibration wizard — the Nextruder loadcell sets Z automatically before every print, as part of mesh bed levelling. Judge B00-P1's first layer by Ellis' smooth-bottom rule: no gaps between beads, no ridging. If it needs a nudge, long-press the knob during the first layer → Live Adjust Z. That nudge is not remembered for the next print, so if you needed one, fix the cause (sheet not seated, debris under it, nozzle not fully seated, glue film uneven) before the next plate rather than nudging every plate. (Prusa KB — Live Adjust Z · Ellis — first layer squish)
4. Print plate B00-P1. It carries the coupons for both gates.
5. Measure with calipers, at mid-height (not across the first layer).

Gate A — no kit needed (the cube)

Coupon	Nominal	Accept	If out of spec
Voron_Design_Cube_v7 X and Y	30.00 mm	±0.15 mm	>+0.15: confirm shrinkage compensation is 0 % and XY compensation is 0. Still over → reduce extrusion multiplier in 1 % steps. Do not dial in negative XY compensation — it wrecks the bearing fits.
Voron_Design_Cube_v7 Z	30.00 mm	±0.10 mm	over → first layer under-squished; under → over-squished (Live Adjust Z, then fix the cause — item 3)
Cube first layer vs mid-height X	—	difference ≤ 0.15 mm	bigger → elephant-foot compensation is wrong; adjust in 0.05 mm steps
Cube corner snap test	—	must not delaminate along a layer line	delamination → chamber too cold or fan too high → drop min/max fan to 0/15 %

The four rows are also a calculator at Step B00.5: type the measurements in, press Check, and it keeps the verdict and shows the advice from this table for whichever row failed.

Corner snap: grip one corner in pliers and bend it off. A pass tears across the layers and leaves a rough, fibrous face; a fail peels cleanly along one layer line and leaves a flat shiny face (verify on bench).

Where the two numbers live. Extrusion multiplier: Filament Settings → Filament. Shrinkage compensation: Filament Settings → Advanced. Both belong to the filament preset inside each project, so a change made here does not travel: re-enter it in every later project as you open it, and write the value on the sheet-edge tape so you know what it should be.

Gate A passed → every batch with no press fit is released: B02, B07, and the cosmetics B08 → B09 → B10. Do not start B01 or B03–B06 on Gate A alone: that is 58.4 h and 763 g of bearing-seat and shaft-bore parts printed against an unverified fit — run Gate B first, the same week.

Gate B — bore and inserts now, rail on kit day

When	Coupon	Nominal	Accept	If out of spec
now	Heatset_Practice	7 × M3×H5 inserts from the KADRICK kit — all seven pockets. Shank must caliper ~4 mm (verify on bench)	insert sits flush to 0.2 mm proud, boss does not bulge > 0.2 mm	bulging → iron too hot or pushed too fast — a technique problem, not a slicer one. Do all seven before touching a real part. Ch 00 Steps 00.14–00.15 are this row, same coupon, same session: the coupon's one use.
now	z_drive_retainer_a 625-2RS pocket (same plate)	16.30 mm, measured off the STL; the concentric lip below it is 14.30 mm	caliper across the pocket reads 16.30 mm ±0.15	over → confirm shrinkage compensation is 0 % and XY compensation is 0, then raise extrusion multiplier 1 %. Under → reduce it 1 %. Never fix it with XY compensation.
kit day	the same pocket, on a real 625-2RS (16 mm OD)	—	bearing presses in with thumb pressure, no rocking	this is the press fit the caliper is standing in for; a fail here reprints the retainer and the cube, and re-passes Gate A
kit day	MGN12_rail_guide on the real MGN12 rail	—	slides on with light finger pressure	very tight → over-extrusion; loose → under-extrusion

The inserts are set by the adult with the X-Tronic iron's stock conical tip; the helper never handles the iron. The LDO brass M3 tip lands with the kit: it is fitted at Ch 00 Step 00.13 and re-checked on the first real insert at Step 02.03. Seven of the KADRICK kit's inserts are consumed here; all 153 kit inserts stay for the build.

The same calculator covers Gate B at Step B00.7, with the two kit-day rows marked optional so they can stay blank until carton 1 is open.

Gate B passed → start B01, then B03–B06. Gate B is Step B00.7; its two early rows take about 15 minutes and unblock 58.4 h of printing, so run them in the same week as Gate A. If Gate B moves the extrusion multiplier, re-print the cube and re-pass Gate A before B01.

If the bore is still loose with shrinkage compensation already at 0 % and XY compensation at 0, raise the extrusion multiplier 1 %, reprint the retainer and the cube, and re-pass Gate A before B01 starts. There is no other lever: negative XY compensation buys this one bore at the cost of every other fit in the machine.

1. Which gate releases which batch:

Batch	Gate	Why
B02, B07	A	no bearing seats, no rail fit; their heat-set bosses get inserts on kit day
B08, B09, B10	A	cosmetic; the skirts are where VFA shows most, but they print on the same GT2 belts as the rest of the run — Galaxy Black's fleck hides the ripple, and GT1.5 lands with the INDX + Gen 2 rebuild
B01, B03, B04, B05, B06	B (bore + inserts, run early)	625-2RS pockets, F695 flange seats, MGN12 carriage pattern, Z-joint bolt pattern

Quality gate before every later batch: look at the last plate pulled off. If any part shows (a) a lifted corner, (b) a delaminated layer, or (c) a bore/boss that failed a test fit — fix that before starting the next plate. Don't print 100 g on top of a known problem.

Orientation & brim

Do not change which face sits on the bed. Every Voron, LDO, Klicky, Nevermore and Clicky-Clack STL ships pre-oriented for support-free printing. Rotating a part about Z (in the plane of the bed) to make it fit is fine — it does not change the layer orientation. Two exceptions to the face rule:

1. XY_cable_chain_bridge-Igus-3mm_backer.stl is a community remix and arrives standing 44 mm tall on a 21 mm-wide footprint, unlike the stock part which lies flat. Check it in the preview before slicing — if it isn't sitting on a flat face, lay it down to match the stock [a]_xy_joint_cable_bridge_2hole.
2. The two Clicky-Clack Hinge-L-* parts. As designed the hinges are on the left and the door opens from the right; mirror both hinge files (2.9.6 only — the 3.0 preview has no mirror tool) if the printer's left side will be against a wall. The handle and latch are symmetric. src

No part in this build exceeds 150 mm in Z. Tallest parts, in order:

Part	Height	Footprint	Aspect	Brim in the project
Handle.stl (Clicky-Clack)	60.0 mm	67.9 × 18.6	tall & narrow	5 mm
Latch.stl (Clicky-Clack)	58.0 mm	25.9 × 14.6	very tall & narrow	5 mm
Hinge-L-sleeve/solid-2X ×4	55.0 mm	20.9 × 19.5	very tall & narrow	5 mm
mount.stl (TFT)	44.8 mm	117 × 67	fine	none
z_motor_mount_a/b ×4	42.0 mm	51 × 30.7	borderline	none — add one by hand only if a print shows lift

The committed projects turn parts about Z wherever it packs the bed better — slicer/geom.py picks the orientation, and the plate diagram in each chapter shows the result (drawn from the committed project itself, so a plate you re-arrange in the GUI and save is redrawn by python3 slicer/build_plates.py --from-3mf). That is the allowed kind of rotation: it never changes which face is on the bed. Example: B08-P1 stands the 182 mm rear_center_skirt_350 front-to-back (turned 90° from the STL) beside an unturned side_fan_support, because side by side as shipped the pair needs 254 mm of X on a 250 mm bed.

These brims are already set per object inside slicer/plates/*.3mf — you do not apply them by hand, and a part that is not on one of these two lists gets no brim even when it shares a plate with one that does. The plate diagram at the top of each Load step draws the brim ring, so the check at every plate is the same: the outline shows on the parts the chapter names, and on nothing else.

Separately, the projects brim the long flat parts — not because they're tall, but because 150–182 mm of ASA on a plate lifts at the ends — and a few smaller flat ones with them: 3 mm brim on rear_center_skirt_350, front_skirt_a/b_350, side_skirt_a/b_350, side_fan_support_x2, keystone_panel, power_inlet_IECGS_1mm, exhaust_cover, V2_Duo_Plenum, Regular_Cartridge, wago_221-415_mount_3by5, exhaust_filter_grill, cob_light_strip_mount_100mm. Brim separation stays at PrusaSlicer's 0.1 mm so it snaps off.

If you ever need to add a brim by hand (a reprint that lifted): right-click the object in the object list → Add settings → Skirt and brim → Brim width, and set it on that object only. Never use the global brim setting (Print Settings → Skirt and brim) — it brims every flat part on the plate, including the skirt fan grills ([a]_fan_grill_*) and clips that must not have one.

Parts with built-in supports to break out, not cut: [a]_stealthburner_main_body, Regular_Cartridge (Nevermore), V2_Duo_Plenum. The skirts are single-shell parts — they have no break-away body.

Colour key

Prefix	Meaning	Our filament
(none)	Primary colour	Prusament ASA Galaxy Black
[a]_	Accent colour	Prusament ASA Blue
[o]_	Opaque — must block light	Galaxy Black (perfect)
[c]_	Clear / translucent	Do not print — LDO supplies the SB LED diffuser in clear PETG
_x# suffix	Quantity required to build the machine	e.g. z_joint_lower_x4.stl → print 4 copies of the file

The _x# number is a quantity, not a hint that the file contains that many bodies. Files confirmed to contain more than one body: [a]_stealthburner_main_body (7 — built-in supports), cw2_captive_pcb_cover (2), usb_adapter_mount (2 — mount base + cover), Leviathan_bracket_set (2 — L+R), bottom_panel_hinge_x2 (2), cob_light_strip_mount_* (2 — the mount is a 2-piece assembly), V2_Duo_Plenum (2 — built-in support), usb_adapter_mount_partial_cover (2 — Nitehawk-SB V2). Clicky-Clack's Hinge-L-sleeve-2X / -solid-2X contain one body each — "2X" means print two.

Three non-Voron parts carry no [a]_ prefix but are printed Blue by choice: Handle.stl (Clicky-Clack), ldo_bestagon_insert.stl, and XY_cable_chain_bridge-Igus-3mm_backer.stl (a remix of the accent [a]_xy_joint_cable_bridge_2hole).

Gen 2 belt upgrade — deferred to the INDX rebuild

Where things stand (2026-09-23). The Core One+ is built and commissioned — commissioning closed 2026-09-13 on Gen 1 (GT2) belts, bed flat to ±0.1 mm on printed Z-stop correction caps. The Gen 1→Gen 2 upgrade kit (Prusa order 1787919456) has shipped. The LDO Voron kit has not: Fabreeko's tracking page still shows the V2.4/Trident batch in manufacture, so realistic delivery is late November to late December 2026.

Decision (2026-09-23): the Gen 2 upgrade is deferred to the INDX 8-tool conversion this winter, not done before B00. Prusa documents the combined install: https://help.prusa3d.com/article/assemblling-the-prusa-indx-core-one-with-the-gen-2-upgrade_1147602. The whole 157.0 h / 22-plate run — B00 → B10, numeric order — prints on the current Gen 1 GT2 belts. This supersedes the earlier "Gen 2 before B00" baseline and its 2026-10-15 contingency; there is no mid-run pause, because nothing changes the machine under the parts.

Why. The printer is commissioned and printing well; the upgrade fixes neither open issue on the bench (loadcell/heater noise, PETG infill strands); a heatbed swap now would put the closed bed calibration (±0.1 mm on the +2.8 front-left Z-stop correction cap) at risk and force a belt retune, an XY-homing recal and a start-G-code re-sync of all 22 committed .3mf projects; GT1.5's benefit is cosmetic (less VFA) and Galaxy Black's metallic fleck already hides it; and one teardown (with INDX) beats two.

Two checks replace it, before B00 — see B00 § Before B00 checks for the full tick list (verify on bench):

1. Belt pluck check. Belt tuning was never formally closed. Pluck mid-belt (motors disabled, head front, X centred) via belt.connect.prusa3d.com or the Prusa app; targets upper ≈96 Hz, lower ≈92 Hz, ≤8 Hz apart. Out of range → Control → Calibrations & Tests → Belt Tuning, tensioner screws half a turn, evenly and alternately.
2. Hot first-layer check, at ASA bed temperature. The bed has only been verified flat at 60 °C, not at the Voron ASA batches' 50–60 °C chamber / hotter bed. Its project is committed: slicer/checks/hot-first-layer.3mf, five 30×30×0.2 mm squares (four corners + centre, one layer) in Galaxy Black ASA on B00-P1's presets, 110 °C bed, cold start G-code. It runs after Step B00.0, whose wizard installs those presets. Pass: no "bed not aligned" prompt, every square 0.17–0.23 mm, spread ≤0.05 mm.

Revised 2026-10-01: the expansion joints and the nozzle wiper go in before B00; only the belts wait. Both parts are fitted in one chapter of Prusa's Gen 2 upgrade guide, 4. Heatbed upgrade (expansion joints, heatbed spacer and aligner, then the nozzle wiper); chapter 3, Belts upgrade, is skipped, and so are the Belt tensioning and Changing the printer edition steps of chapter 5. How much of chapter 2's disassembly the heatbed needs with the Nextruder left on is (verify on bench). Buddy 6.8.1 keeps separate switches "for printers that have been retrofitted with only some of the accessories": Settings → Hardware → Expansion Joints Gen 2 and → Nozzle Wiper go on, GT1.5 Belts stays off. Edition = CORE One+ Gen2 would also set GT1.5 steps/mm on GT2 belts. Taking the heatbed off reopens the closed ±0.1 mm bed, so the six-point bed check and the Z-stop correction caps are redone before the hot check. The hot check then runs twice, on the cold start and on the heat-then-off start (slicer/checks/hot-first-layer-heat-off.3mf), and the better one goes into all 22 plates with slicer/sync_start_gcode.py. With the wiper on, G29 P9 heats the nozzle itself, taps a touchpoint beside the brush with the heater on, and leaves a hold target behind; the heat-then-off start switches the heater off again before the mesh for that reason.

After the eventual INDX + GT1.5 belt rebuild, before restarting prints:

1. Firmware ≥ 6.8.1 (carries GT1.5 belt support and the Gen 2 expansion joints; the combined INDX + Gen 2 guide supersedes the standalone Gen 2 KB article once both go in together).
2. Re-tension both belts and re-square the gantry.
3. Re-run the self-test and input shaper calibration.
4. Re-print Voron_Design_Cube_v7 and re-pass Gate A — steps/mm changes with the pulleys, so the dimensional gate must be re-passed before you print 399 g of skirts (B08) or anything else on the rebuilt machine. Judge the cube's first layer as in item 3 of the calibration sequence — the loadcell re-zeroes on its own; there is no wizard to redo.

The same rule generalises: any toolchain change — slicer version (2.9.6 → 3.0), profile bundle, belts, pulleys or nozzle — re-runs Gate A (and Gate B, if run before the kit's fits are all confirmed) before the next plate. For this build that next applies at the INDX + Gen 2 rebuild, not mid-run.

Do not try to interleave a belt/tool upgrade with a running plate — the Nextruder and bed have to come apart.

STL source of truth

Reference only — you need this only if you re-derive the plates. slicer/fetch_stls.py fetches these exact trees and slicer/stl/MANIFEST.sha256 pins every file to a commit; printing from the committed projects needs none of it.

Do not use a re-hosted zip or a Printables mirror. The Voron repos have no release tags — the manifest records the branch head so a mid-build reprint is identical to the first print.

#	What	Repo / branch	Path
1	Voron 2.4 R2 core	VoronDesign/Voron-2 branch Voron2.4	/STLs
2	Stealthburner + Clockwork 2	VoronDesign/Voron-Stealthburner branch main	/STLs
3	LDO kit-specific	MotorDynamicsLab/LDOVoron2 branch main	/STLs
4	LDO touchscreen mount	MotorDynamicsLab/LDOVoronTrident branch master	/STLs/BTT Pi TFT4.3 Mount
5	Nitehawk-SB	MotorDynamicsLab/Nitehawk-SB branch master	/STLs
5b	Nitehawk-SB V2 (Rev D+ USB cover)	MotorDynamicsLab/Nitehawk-SB-V2 branch master	/STLs
6	Klicky probe	jlas1/Klicky-Probe branch main	/Probes/KlickyProbe/STL + /Printers/Voron/v1.8_v2.4_Legacy_Trident/v1.8_v2.4_Legacy_Trident_STL
7	Nevermore Micro V5 Duo	nevermore3d/Nevermore_Micro branch master	/V5_Duo/V2
8	Clicky-Clack door	tanaes/whopping_Voron_mods branch main	/clickyclacky_door/STLs
9	Backer-aware cable bridge (conditional)	tanaes/whopping_Voron_mods branch main	/extrusion_backers/STLs

PrusaSlicer 3.0 preview

Reference only — first build: 2.9.6 for everything.

3.0.0-alpha12 (2026-09-21, a bug-fix build over alpha11) is usable for the cosmetic batches but changes enough that it is not a drop-in. Verified against the 3.0 branch, not recalled:

• Profiles moved from the single PrusaResearch.ini to per-preset YAML under resources/presets/prusa-research-fff/PrusaResearch/.
• ASA shrinkage is still 0.22 % on CORE One (preset-filament-common.yaml, id *shrinkage_ASA*, condition: printer.base_model=~/(COREONE|COREONE_INDX)/) — so the zeroing override in the Filament table above is unchanged and just as important.
• 0.20mm STRUCTURAL @COREONE 0.4 exists under the same name. Its bottom_solid_layers default is 3 in 3.0, not 4 — the action is unchanged: set it to 5.
• support_material is now one enum (CORE One base value enforcers_only) replacing the 2.x support_material + support_material_auto pair. Set it to none.
• The printer is picked as the model "Prusa CORE One & CORE One+" with the nozzle chosen per-tool, not as a …0.4 nozzle preset.
• Extrusion-width auto/percent values now resolve against nozzle diameter instead of layer height. Harmless here — every width in the override table is an explicit mm value.
• No .ini import. 3.0 has no Import Config / Import Config Bundle / Export Config, so slicer/voron-coreone-asa.ini does not cross the boundary. What does cross is the plate projects: 3.0 imports a 2.x 3MF project config and maps it to system presets, so open slicer/plates/<plate>.3mf and check every override by hand against the tables above. A 3.0 project opened in 2.9.6 loads geometry only — the configuration is discarded (2.9.1+ warns).
• No mirroring. Mirroring is still on the 3.0 queue, so the B10 Hinge-L-* mirror cannot be done in the preview: slice B10 in 2.9.6, or mirror the STL outside the slicer.