3D Printing Info
CoreXYELEGOO

Elegoo Centauri Carbon 2

Open in interactive comparison Browse printers

140 catalog fields126 catalog sources

Architecture referenceCoreXY architecture illustration; not an image of this model
CoreXY · Generic illustration, not this model’s appearance or construction. View product photos
Start here

At a glance

Explore this printer

Build volume

Advertised 256 × 256 × 256 mm; platform 260 × 260 mm. Firmware agrees exactly: x_range=0,256, y_range=0,256, z_range=0,256. Axis travel exceeds the print area: X position_min=-2.1 → position_max=256.5; Y -2.1 → 270.0; Z -10 → 258.0. Slicer printable_area 256×256, printable_height 256, with a bed_exclude_area of 246×0 → 256×20 mm

Advertised dimensions; check usable space and variants.

Sources & dates
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L11–13, L41–42, L67–68, L94–95; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json · 2026-08 · local

Nozzle temperature

Claimed 350 °C. Firmware limit max_temp=365 in the shipping config (printer_dsp.cfg L134), with sensor_type=NTC R25 350 and min_temp=-100. The non-shipping printer.cfg in the same repo sets max_temp=600 with sensor_type=NTC 100K beta 4300 — this drew a public GitHub issue

Published limits; hardware and firmware limits may differ.

Sources & dates

Price snapshot

$379.00 USD displayed sale price on us.elegoo.com (badged −16 % against $449.00 list) as of 2026-08-19. The same page's machine-selector chip and the Shopify products.json variant both read $399.00 (compare-at $449.00). Certified pre-owned: $375.99. Amazon B0G4TPZPZM: $449 per Tom's price widget (Feb 2026) and $500 per Notebookcheck (Mar 2026)

Dated listing, not a live offer. Read the configuration and region.

Sources & dates
A closer look

Hardware, software & ownership

Full specifications inside each chapter

Understand the hardware, explore compatible changes and find a diagnostic starting point when something goes wrong. Expand the specification sections for every recorded value, condition and original reference.

Machine & motion

Dimensions, construction and the hardware that moves the print.

Model, variants & pricing6 fields
Elegoo Centauri Carbon 2 — Identity & Market
Property and unitPublished value and conditionsSources and dates
Manufacturer & exact marketed model name (verify the exact product identity first)ELEGOO Centauri Carbon 2 Combo. Internal ids: slicer model_id: "Elegoo-CC2", printer_model: Elegoo Centauri Carbon 2, machine setting_id: ECC204 (0.4 nozzle), filament suffix @ECC2; firmware release folder printer-realase-2602101855; mainboard part number EF-M5-V1.0.073. Manual PDF title block: 11.201.0124B Centauri Carbon 2 Combo安装说明书-EN-V1.1
2 sources
Release date (announcement and actual shipping)Announced 2026-01-19; on sale 2026-01-26, 09:00 EST / 17:00 CET.
4 sources
Launch MSRP (USD)US $449.00. Regional launch MSRPs: €439, £339, C$619, A$999, ¥79,999
2 sources
Street price (USD; dated snapshot per entry)$379.00 USD displayed sale price on us.elegoo.com (badged −16 % against $449.00 list) as of 2026-08-19. The same page's machine-selector chip and the Shopify products.json variant both read $399.00 (compare-at $449.00). Certified pre-owned: $375.99. Amazon B0G4TPZPZM: $449 per Tom's price widget (Feb 2026) and $500 per Notebookcheck (Mar 2026)
4 sources
Variants & sibling models — and how to tell them apart (spec-contamination risk)Five live Centauri siblings on us.elegoo.com (2026-08-19): (1) Centauri Carbon 2 Combo $399/$379 — printer + CANVAS; 500×480×743 mm, 19.35 kg. (2) Centauri Carbon 2 (non-Combo) $349 — *identical printer, no CANVAS*; 398×404×490 mm, 17.5 kg. (3) Centauri 2 Combo $329. (4) Centauri 2 $279. (5) Centauri Carbon (CC1) $419.99. Plus ELEGOO × emoji® CC2 Combo $449 (cosmetic edition). CANVAS standalone: $150 (with cover) / $120 (without); CC2-non-Combo + CANVAS-no-cover bundle $389
3 sources
Market position: superseded? successor announced?Current flagship of the Centauri line; no CC3 or CC2 successor announced. Elegoo has widened the line *below* it (Centauri 2 / Centauri 2 Combo, Jun–Jul 2026) and continues shipping CC2 OTA firmware (latest catalogued v02.00.02.00, 2026-05-28). Benchmarked by reviewers against Bambu Lab P1S Combo ($549), Bambu P2S Combo ($799), Creality K2 Combo ($549), Creality SPARKX i7 ($399) and Anycubic Kobra X
2 sources
Size, frame & enclosure6 fields
Elegoo Centauri Carbon 2 — Physical & Frame
Property and unitPublished value and conditionsSources and dates
Build volume X×Y×Z (mm) — advertised vs actually usableAdvertised 256 × 256 × 256 mm; platform 260 × 260 mm. Firmware agrees exactly: x_range=0,256, y_range=0,256, z_range=0,256. Axis travel exceeds the print area: X position_min=-2.1 → position_max=256.5; Y -2.1 → 270.0; Z -10 → 258.0. Slicer printable_area 256×256, printable_height 256, with a bed_exclude_area of 246×0 → 256×20 mm
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L11–13, L41–42, L67–68, L94–95; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json · 2026-08 · local
Machine footprint (mm) & clearance needed (cable/spool overhang)Combo: 500 × 480 × 743 mm (= 19.68 × 18.89 × 29.25 in). Bare CC2: 398 × 404 × 490 mm. Package: Combo 490 × 495 × 560 mm (gross 23.8 kg); bare CC2 465 × 460 × 560 mm; CC1 480 × 475 × 560 mm. Extra clearance needed to the right for four side-mounted spool holders and above for the tophat
1 source
  • evidence/elegoo_official_product_page.md; docs/CC2_manual_EN.pdf p.01; tomshardware.com; voxelmatters.com; evidence/elegoo_cc2_vs_cc1_official_comparison.md · 2026-08 · local
Net weight (kg)Combo: 19.35 kg (gross 23.8 kg). Bare CC2: 17.5 kg (gross 20.4 kg)
1 source
  • evidence/elegoo_official_product_page.md; docs/CC2_manual_EN.pdf p.01; tomshardware.com ("19.35 kg / 42.65 lbs"); techradar.com body ("42.66 lb"); notebookcheck.net ("abo · 2026-08 · local
Frame construction & materialIntegrated die-cast aluminium frame and chassis, SPCC steel panels, injection-moulded covers; metal side panels and a glass front door
1 source
Enclosure: open / enclosed; panel material; door/lidFully enclosed. Metal side panels + glass front door + removable "tophat" lid (acrylic/Perspex) that also covers the CANVAS module; a thermal cover ships in the box. Rear carries a servo-actuated exhaust grille and an air filter
1 source
  • tomshardware.com ("fully enclosed … glass front door … a very large 'top hat'"); techradar.com ("Perspex top … essentially making it almost airtight"); evidence/elegoo_of · 2026-08 · local
Passive chamber temperature reached (if enclosed; measured values preferred)MEASURED: peak ~47 °C, sustained ~45 °C, from a 17 °C ambient, with the tall cover fitted. Aurora Tech placed an independent thermometer inside the chamber during a standard-ABS print in a 17 °C garage: the internal temperature reaches ~35 °C shortly after the print starts, exceeds 40 °C after about an hour, peaks at ~47 °C, then settles at ~45 °C for the remainder as the part grows and the bed descends. On a separate, smaller PA-CF job the enclosure read ~41 °C near the end. Aurora Tech is based in Danville, California, so this was obtained on 110-120 V mains, i.e. on the ~250 W bed (see I2), which makes the figure directly applicable to US buyers. Firmware-verified surrogate, consistent with the measurement: a chamber thermistor exists ([ztemperature_sensor box], NTC 100 K B3950, max_temp=110, over-temp cutoff ot_box=100) and the chamber fan is thermostatically staged at T0=30, T1=38, T2=37, T3=40, T4=45 C. An owner reports being able to request a slicer-side chamber 50 C target / 45 C minimum
2 sources
Motion system8 fields
Elegoo Centauri Carbon 2 — Kinematics & Motion Hardware
Property and unitPublished value and conditionsSources and dates
Kinematics (bedslinger i3 / CoreXY / other)CoreXY — kinematics=corexy. Fixed gantry; bed moves in Z
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L6; corroborated by tomshardware.com, techradar.com, notebookcheck.net · 2026-08 · local
X-axis guide type (rods / linear rails / V-wheels; size)Ø8 mm hardened smooth rods with LM8LUU / LMS8LUU linear ball bearings. Not linear rails, not V-wheels
1 source
Y-axis guide typeSame as X: Ø8 mm smooth rods + LM8LUU/LMS8LUU linear bearings
1 source
  • docs/CC2_BOM.pdf (Y光轴, Y光轴固定钣金, Y轴光轴固定块, Y轴轴承固定钣金); tomshardware.com · 2026-08 · local
Z-axis: leadscrew count & type, sync mechanism, anti-backlash3 lead screws + 3 linear guide rods, all driven from a single Z stepper via a belt. Firmware declares exactly one [stepper_z]: rotation_distance=8 (8 mm Z per motor rev), run_current=1.0, position_min=-10, position_max=258. Z endstop is a physical optical limit switch (endstop_pin=^PB10), *plus* the load-cell probe for Z-zero. No anti-backlash nut documented
1 source
Belts (type, width, e.g. Gates?)2M-pitch (2 mm), 6 mm-wide open-ended timing belt — BOM part 2M-W6开口同步带 — with dedicated XY belt-tensioning springs (X/Y轴同步带张紧弹簧), tensioner-idler brackets (XY轴皮带张紧惰轮支架) and belt clamps (XY皮带固定卡片). No Gates branding in any source. Firmware-consistent: rotation_distance=39.872 on X and Y
1 source
  • docs/CC2_BOM.pdf; firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L38, L64 · 2026-08 · local
Stepper motors (frame size, step angle, brand if known)X and Y: two 42 × 60 mm NEMA 17 (BOM 42-60步进电机; Elegoo: "Dual 4260 stepper motors independently drive the X and Y axes"). Extruder: round 36 × 20 mm NEMA-14-class, P/N BJY36D12-04V28, Shenzhen Keli Motor Ltd (BOM 36x20圆形步进电机). CANVAS: 4 × permanent-magnet steppers with worm gearboxes, Shenzhen Wanzhida Motor. Step angle: full_steps_per_rotation is absent from the config → Klipper default 200 steps/rev = 1.8°
3 sources
Stepper drivers (exact model e.g. TMC2209; UART/SPI; sensorless homing?)TMC2209 throughout, UART, with sensorless homing on X and Y. X: uart_pin=PE6, tx_pin=PE7, run_current=1.2, hold_current=0.8, driver_SGTHRS=130, diag_pin=PF0, endstop_pin=tmc2209_stepper_x:virtual_endstop. Y: same currents, SGTHRS=130, diag_pin=PF3. Z: run_current=1.0, SGTHRS=120, but uses the physical endstop. Extruder (on toolhead MCU): uart_pin=toolhead:PB11, uart_address=2, run_current=0.8, hold_current=0.2. sense_resistor=0.1 throughout. CANVAS uses 4 × AT8833 (DRV8833 clone) — not TMC
2 sources
Microstepping / interpolation settings (firmware evidence)Microsteps=16 on every axis, with interpolate=True on every TMC2209 (1/16 commanded, interpolated to 1/256 in-driver)
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L37, L51 (X); L63, L77 (Y); L89, L104 (Z); L119, L148 (E) · 2026-08 · local
Speed & acceleration10 fields
Elegoo Centauri Carbon 2 — Speed & Acceleration
Property and unitPublished value and conditionsSources and dates
Max print speed (claimed, mm/s)≤500 mm/s
1 source
  • evidence/elegoo_official_product_page.md; docs/CC2_manual_EN.pdf p.01; tomshardware.com; techradar.com; notebookcheck.net; voxelmatters.com · 2026-08 · local
Recommended/typical print speedElegoo's own spec table says "Recommended 250 mm/s". Reviewer practice: Tom's ran 160 mm/s avg (multicolour PLA), 150 mm/s (PETG), 100 mm/s (PAHT-CF and PC), 50 mm/s (TPU). 3DwithUS used 300 mm/s inner / 250 mm/s outer walls
1 source
Max travel speed500 mm/s — max_velocity=500 is a global ceiling covering travel; slicer sets machine_max_acceleration_travel=20000. Z far slower: max_z_velocity=50
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L7, L9; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json · 2026-08 · local
Max acceleration (claimed, mm/s²)20 000 mm/s² maximum; 10 000 mm/s² "Default Acceleration"
1 source
  • evidence/elegoo_official_product_page.md (both rows present); evidence/elegoo_cc2_vs_cc1_official_comparison.md · 2026-08 · local
Firmware velocity limit (`max_velocity` or Marlin equivalent — exact value)Max_velocity= 500; max_z_velocity= 50
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L7, L9 · 2026-08 · local
Firmware acceleration limit (`max_accel` / `DEFAULT_MAX_ACCELERATION` — exact value)Max_accel= 20000; max_z_accel= 1200. Slicer: machine_max_acceleration_extruding=20000, machine_max_acceleration_z=500. Start g-code clamps outer walls with M204 S{min(20000, max(1000, outer_wall_acceleration))}
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L8, L10; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json (incl. machine_start_gcode) · 2026-08 · local
Jerk / square-corner-velocity setting (exact value)Square_corner_velocity=9. (9 mm/s, ~1.8× Klipper's 5.0 default). Slicer mirrors it: machine_max_jerk_x = machine_max_jerk_y = 9, machine_max_jerk_e = 1
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L14; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json · 2026-08 · local
Max volumetric flow claimed (mm³/s)The vendor slicer profiles are the operative claim — PLA 21, ABS 18, PC 16, PLA-CF 16, PETG-CF 12, PETG 11, PET-CF 8, PAHT-CF 6, TPU 95A 3.2 mm³/s. Elegoo publishes no headline flow figure
1 source
  • firmware/ElegooSlicer_ECC2/Elegoo_{PLA,ABS,PC,PLA-CF,PETG-CF,PETG,PET-CF,PAHT-CF,TPU_95A}_@ECC2.json (filament_max_volumetric_speed) · 2026-08 · local
Max volumetric flow independently measured (mm³/s, method)Not published
1 source
  • — · 2026-08 · local
Real-world speed benchmarks (e.g. Benchy time from third-party tests)No rigorously-specified Benchy exists for the CC2. Actual data points: CreativeBloq ~40 min for a colour-split multi-part Benchy; 3DwithUS ~2 h for a 4-colour Benchy and ~35 min for the built-in single-colour "Little Buddha"; VoxelMatters "just over 20 minutes" for the *same* Buddha; Tom's 4 h 55 min / 7 Baby Chameleons @0.2 mm, 160 mm/s avg; 4 h 20 min PETG Slinky @0.2 mm, 150 mm/s; 6 h 06 min / 7 PC utility knives @0.2 mm, 100 mm/s; 2 h 12 min PAHT-CF part @0.2 mm, 100 mm/s. Aggregator smith3d quotes "~31 min at 500 mm/s / ~48 min at 300 mm/s" and an "18 min 19 s speed-optimised Benchy on the Centauri platform" — provenance unverifiable, and "the Centauri platform" may mean CC1
4 sources

Printing system

Extrusion, heating, leveling and material handling.

Vibration compensation5 fields
Elegoo Centauri Carbon 2 — Vibration Compensation / Input Shaping
Property and unitPublished value and conditionsSources and dates
Input shaping supported? Implementation (Klipper IS / Marlin ZV / proprietary)Yes — stock Klipper-style [input_shaper] + [resonance_tester] in Elegoo's C++ Klipper re-implementation. Marketed as "Active Vibration Compensation"
1 source
Accelerometer: present? model (ADXL345 / LIS2DW12 / …), location(s), countOne accelerometer: a single [lis2dw] on the toolhead, software-SPI (cs_pin=toolhead:PA4, sclk=PA5, mosi=PA7, miso=PA6, 5 MHz, axes_map=x,y,z). Teardown identifies it as LIS2DW12 on SPI1 with labelled test points CS/SCLK/MOSI/MISO = PA4/PA5/PA7/PA6 — an exact pin-for-pin match to the config
1 source
Calibration mode: automatic per-print / user-run routine / factory-fixed valuesUser/self-check-run routine that stores results — NOT per print. [resonance_tester] carries shaper_calibrate_gcode=G28\n, i.e. a discrete calibration action. The per-print machine_start_gcode contains BED_MESH_CALIBRATE but no shaper calibration. Notebookcheck: "the one click calibration takes care of optimizing the nozzle, bed level, and vibration compensation. The whole calibration process takes about 26 to 27 minutes"
1 source
Shaper type & default frequencies (firmware evidence preferred)Shaper_type_x = mzv, shaper_freq_x = 53.6; shaper_type_y = mzv, shaper_freq_y = 44.8; damping_ratio_x = damping_ratio_y = 0.1. Search space: min_freq=5, max_freq=90, accel_per_hz=100, hz_per_sec=2, max_smoothing=1, probe point 128,128,5
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L277–282, L292–300 · 2026-08 · local
Effectiveness evidence (ringing/VFA tests from reviews). TechRadar's structured test scores XY resonance 2.5 / 5 and Z-axis alignment 2.5 / 5 — by far the weakest two of its eight sub-scores. Against that, Tom's saw "no color bleeding using the default purge values" on a 7-part multicolour plate; CreativeBloq: "layer lines are barely visible, corners and edges are sharp" and "there were very few of those annoying wispy bits of filament you can get when the nozzle lifts off"; 3DwithUS: "All lettering came out sharp and clearly defined"
1 source
Extrusion pressure control5 fields
Elegoo Centauri Carbon 2 — Extrusion Pressure Control
Property and unitPublished value and conditionsSources and dates
Pressure advance (Klipper) / linear advance (Marlin) supported?Yes — Klipper-style pressure advance
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L135; firmware/ElegooSlicer_ECC2/*_@ECC2.json (enable_pressure_advance: ['1']) · 2026-08 · local
Enabled by default? Factory value? (firmware evidence)Yes. Firmware default pressure_advance=0.04. The slicer then overrides it per material: PLA/ABS 0.04, PC 0.052, PAHT-CF 0.052, TPU 95A 0.1
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L135; firmware/ElegooSlicer_ECC2/Elegoo_ABS_@ECC2.json, Elegoo_PC_@ECC2.json, Elegoo_PAHT-CF_@ECC2.json, Elegoo_TPU_95A · 2026-08 · local
Auto-calibration for PA/LA? How does it work?Marketed, but no per-print PA routine exists in the shipped g-code. machine_start_gcode contains no PA calibration call; there is no PA macro in [g180]. Elegoo's product page nonetheless advertises "Pressure Advance Calibration" as a headline feature
1 source
Any physical nozzle pressure / load-cell extrusion sensing?No. A load cell exists ([load_cell_probe], CS1237 24-bit ADC on the bed MCU, 4-wire Wheatstone strain-gauge bridge) but it is wired to the bed as the Z probe — it is not read for extrusion force. Filament *presence* and *motion* are sensed optically/by encoder instead
1 source
Related settings (PA smooth time etc., firmware values)Pressure_advance_smooth_time=0.04 (Klipper default is 0.040 — unchanged). Extruder also sets instantaneous_corner_velocity=1, max_extrude_only_distance=300, max_extrude_only_velocity=60, max_extrude_only_accel=5000, max_extrude_cross_section=9999999, and retract_restart_extra_toolchange=['0.5'] in the slicer
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L136–142; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json · 2026-08 · local
Extruder6 fields
Elegoo Centauri Carbon 2 — Extruder
Property and unitPublished value and conditionsSources and dates
Type (direct drive / bowden) & design nameDirect drive, "Dual-Gear Direct Extruder (1 Fan)" (Elegoo's own wording). On the Combo the extruder additionally carries a filament multiplexer (4-into-1 hub) and a filament cutter
2 sources
Gear ratio & drive gear materialGear_ratio=52:10 (5.2 : 1), rotation_distance=28.8. Hobbed gear: 10 mm nominal diameter, SDK11 tool steel
1 source
Filament diameter1.75 mm
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L123 (filament_diameter=1.75); evidence/elegoo_official_product_page.md; docs/CC2_manual_EN.pdf · 2026-08 · local
Filament path (PTFE-lined? all-metal? reverse bowden?)Combo: four PTFE Bowden tubes run from CANVAS on top of the machine to a 4-in-1 filament hub / multiplexer screwed to the toolhead; from the multiplexer a 4 mm OD metal tube feeds the extruder, replacing the reverse-Bowden PTFE. Melt zone is all-metal with a titanium-alloy heat break — no PTFE at the melt zone. Non-Combo: conventional reverse-Bowden PTFE
1 source
Claimed extrusion force / grip (if published)Elegoo publishes no extrusion-force figure
1 source
  • Searched: evidence/elegoo_official_product_page.md, docs/CC2_manual_EN.pdf, evidence/elegoo_cc2_vs_cc1_official_comparison.md, all eight reviews, and OpenCentauri's toolh · 2026-08 · local
Auto load/unload routineYes, fully automatic in both directions. [filament_load_unload] defines two-stage load and unload macros; the load routine heats to 250 °C (M104 S250 / M109 S250), parks at the purge station via G180 S800, then runs G180 S812 → S811 → S810 → S800; unload runs G180 S813 (cutter position), retracts E-0.8, then E-100 F600. CANVAS auto-grabs filament on insertion; Elegoo lists "Auto Filament Unloading: Yes"
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L431–436 ([filament_load_unload]); L344–366 ([canvas_dev] pre_load_fila_dist=150, load_fila_dist=1000, unload_fila_dist · 2026-08 · local
Problems & troubleshooting

Nozzle blockage: inspect wear, then test extrusion

ELEGOO’s shared guide starts by checking nozzle-tip wear, then uses a controlled extrusion test, cleaning needle and cold-pull route. Follow the illustrated toolhead reassembly and test before returning to a long print.

Applies to: Centauri 2/Combo and Centauri Carbon 2/Combo are explicitly named; use the matching toolhead and temperature instructions.

Official troubleshooting Sources & dates
Hardware & upgrades

FlowTech with the CC2 thermistor variant

This replacement changes to the FlowTech cold-change nozzle system. Select the Centauri Carbon 2 option because the manufacturer distinguishes its thermistor parameters from the first-generation printer.

Applies to: Centauri Carbon 2/Combo; use its CC2 option. The listing does not establish compatibility with the open-frame Centauri 2.

Manufacturer documentation Sources & dates
Hotend & nozzle10 fields
Elegoo Centauri Carbon 2 — Hotend & Nozzle
Property and unitPublished value and conditionsSources and dates
Hotend design (proprietary/standard; quick-swap mechanism?)Proprietary Elegoo hotend, shared design lineage with CC1 but with a higher-temperature heater. The hotend sits on a main toolhead PCB plus a separate "supplementary board" (2×4 pin) that breaks out thermistor, heater and hotend fan. No tool-less quick-swap mechanism is documented; the nozzle is a separate serviceable part
2 sources
Max nozzle temp claimed AND firmware max temp limit (both values)Claimed 350 °C. Firmware limit max_temp=365 in the shipping config (printer_dsp.cfg L134), with sensor_type=NTC R25 350 and min_temp=-100. The non-shipping printer.cfg in the same repo sets max_temp=600 with sensor_type=NTC 100K beta 4300 — this drew a public GitHub issue
1 source
Heater cartridge power (W)60 W ceramic plate-type PTC heater, ~9.6 Ω (24 V → 24²/9.6 = 60 W, self-consistent)
1 source
Temperature sensor type (thermistor model / thermocouple) & ratingFirmware: sensor_type=NTC R25 350 (an Elegoo-defined curve; the name implies R25 characterisation to 350 °C), on sensor_pin=toolhead:PA3. Teardown: "Glass bead NTC-200k", beta unknown ("likely 3950 or 4300"). The non-shipping printer.cfg uses NTC 100K beta 4300. Bed and chamber both use NTC 100 K β3950
1 source
Stock nozzle size & sizes included in box0.4 mm fitted; only 0.4 mm supplied. No spare nozzle in the box (the box does include a "Spare Nozzle Wiper" and an unclogging pin, which is a different thing)
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L122 (nozzle_diameter=0.4); evidence/elegoo_official_product_page.md (spec "Nozzle Diameter 0.4 mm"; "What's in the Box · 2026-08 · local
Stock nozzle material (brass / hardened steel / plated copper / composite tip)Bimetal: brass body with a hardened-steel tip. Elegoo's own part description: "0.4 mm nozzles with brass bodies for fast, even heating and hardened steel tips for long-lasting wear resistance". BOM part 黄铜硬化钢喷嘴 (brass hardened-steel nozzle)
1 source
Nozzle mount standard (V6 / MK8 / Unicorn / proprietary) & swap procedureProprietary Elegoo Centauri-series nozzle (fits the whole Centauri family — CC1, CC2, Centauri 2 — per the part name "for Centauri Series"). Not V6/MK8/Unicorn. No documented tool-less swap; the manual's maintenance guidance is limited to clearing residue with the supplied unclogging pin
2 sources
Available nozzle sizes/materials (official + third-party compatibility)Official: 0.2 / 0.4 / 0.6 / 0.8 mm, all brass-body + hardened-steel-tip. "Multi-Size Brass Hardened Steel Nozzle Kit for Centauri Series" ($24.99) contains all four; single 0.4 mm kit $22.99. Slicer ships full profile sets for all four sizes (nozzle_diameter: "0.4;0.2;0.6;0.8", 21 process profiles spanning 0.08–0.48 mm layers). Third-party: none documented
1 source
All-metal hotend? PTFE near melt zone?Yes, all-metal, with a titanium-alloy heat break. No PTFE at the melt zone — PTFE stops at the multiplexer/hub above the extruder
2 sources
Melt-zone / high-flow design details (CHT-style, length)In detail. Known: plate-type ceramic PTC heater (60 W), titanium heat break, 3010 axial heat-break fan (12 000 rpm, 0.10 A @24 V), silicone sock, bimetal nozzle. No melt-zone length, bore diameter, or CHT/high-flow geometry is published or measured anywhere
1 source
Build plate & heating6 fields
Elegoo Centauri Carbon 2 — Heated Bed & Surface
Property and unitPublished value and conditionsSources and dates
Max bed temp claimed AND firmware limitClaimed 110 °C (at 25 °C ambient). Firmware max_temp=125, with a separate hardware over-temp parameter ot_bed=130. min_temp=-100
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L163–165; evidence/elegoo_official_product_page.md; docs/CC2_manual_EN.pdf p.01 · 2026-08 · local
Bed heater type (DC/AC mains) & wattageAC mains-powered heater switched by a solid-state relay (SSR) board — NOT a 24 V DC bed. Measured resistance ~48.4 Ω; 1000 W @ 220 V / 250 W @ 110 V. Safety: ground present + thermal fuse. Control is a PWM signal (heater_pin=PG13, pwm_cycle_time=0.016 ≈ 62.5 Hz) into the SSR; the mainboard's HBED connector carries only the control signal
3 sources
Heat-up time to 60°C (measured preferred)For the CC2. No reviewer timed it
1 source
  • Searched all eight CC2 reviews (Tom's, TechRadar, Notebookcheck, Hackster, GamesRadar, 3DwithUS, CreativeBloq, VoxelMatters) plus general web search · 2026-08 · local
Bed surface (material, texture, coating), single/double-sidedDual-sided spring-steel flex plate: textured PEI on one side, a "PLA-Specific" smooth surface on the other. Plate stack: 3 mm aluminium plate + 1.5 mm magnetic sheet. Platform 260 × 260 mm
1 source
Magnetic removable plate? Spare plate availability & costYes, magnetic removable. Spares: "Dual-Sided Build Plate for Centauri Series" $22.99; 3-pack $39.99; an older revision "(Original)" $21.99
2 sources
Bed thermal uniformity (third-party thermal imaging if available)Not published
1 source
  • Searched all eight CC2 reviews plus targeted web searches for thermal-camera/FLIR bed-uniformity tests of the CC2; also checked My Tech Fun's bed-heating comparison work · 2026-08 · local
Leveling & Z-offset6 fields
Elegoo Centauri Carbon 2 — Leveling & Z-Offset
Property and unitPublished value and conditionsSources and dates
Auto bed leveling: probe technology (inductive / strain gauge / load cell / touch probe)Load-cell / strain-gauge nozzle-touch probe in the bed. [load_cell_probe] with sensor_type=cs1237 (CS1237 24-bit ADC) on the dedicated bed MCU (sclk_pin=bed_sensor:PA5, dout_pin=bed_sensor:PA4, gain=A-128, sample_rate=1280). Calibration: counts_per_gram=25.0, trigger_force=200.0 g, safety_limit=1200.0 g, trigger_count=5. Bed hardware: 4-wire Wheatstone strain-gauge bridge with a single amplifier IC
1 source
Mesh grid size (probe point count — firmware evidence)Probe_count=11,11 → 121 points, mesh_min=6,6, mesh_max=246,246, mesh_pps=3,3, algorithm=bicubic, bicubic_tension=0.2, mesh travel speed=500.0, horizontal_move_z=3.0, calibrate_accel=20000, move_check_distance=1.0, split_delta_z=0.01. Slicer mirrors: bed_mesh_min 10,10, bed_mesh_max 243,245, bed_mesh_probe_distance 22,22
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L589–604; firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json · 2026-08 · local
Auto Z-offset (nozzle-touch etc.) or manual (paper method)?Fully automatic nozzle-touch Auto Z-offset — no paper method. z_offset=0 in the probe section (the load cell defines Z=0 by contact). Additionally a nozzle-thermal-expansion compensation table exists: [z_compensation] with nozzle_temp_start=100, nozzle_temp_end=250, nozzle_temp_diff=10, fix_ref_temperature=140.0, and a 16-entry measured offset table nozzle_calibrate_datas = -0.039173 … 0.072890 mm
1 source
Manual leveling screws/knobs present?None documented. The manual's assembly chapter has no bed-levelling-screw step; the bed ships locked by three transport screws that are removed, not adjusted
1 source
  • docs/CC2_manual_EN.pdf (assembly section — "remove the foam … marked in blue to unlock the heated bed"; no levelling-screw procedure); tomshardware.com ("After unpacking · 2026-08 · local
Gantry-leveling / Z-tilt routine?No — and structurally impossible in software. There is one Z stepper driving three belted lead screws, so there is no per-screw actuator to tilt. No [z_tilt] or [quad_gantry_level] section exists in either config
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg (complete section list contains no [z_tilt]/[quad_gantry_level]); single [stepper_z] at L85; tomshardware.com ("three l · 2026-08 · local
Leveling workflow (forced per print? saved mesh? adaptive mesh?)Mesh is measured on every print by default. machine_start_gcode unconditionally issues BED_MESH_CALIBRATE mesh_min=… mesh_max=… ALGORITHM=… PROBE_COUNT=… ADAPTIVE=0 ADAPTIVE_MARGIN=0 FROM_SLICER=1 before G28 and heating. Adaptive meshing exists but is off in the default profile (ADAPTIVE=0); [g180] macros S8/S910 implement BED_MESH_PROFILE LOAD=ADAPTIVE. A saved default profile also exists (G180 S7 → BED_MESH_PROFILE LOAD=default). Elegoo additionally lists "Partial Bed-Leveling" as a CC2-only feature vs CC1. Tom's: "runs a full calibration … during the device self-check. The machine also runs through an optional leveling calibration before each print"
1 source
  • firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json (machine_start_gcode); firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L644+ ([g180] S7/S8/S910/S91 · 2026-08 · local
Filament handling6 fields
Elegoo Centauri Carbon 2 — Filament Handling & Detection
Property and unitPublished value and conditionsSources and dates
Filament runout sensor (type, location)Two-layer detection. (a) Toolhead optical sensor on the filament-detector board — [filament_switch_sensor filament_sensor] switch_pin=!toolhead:PB1, pause_on_runout=true, event_delay=10.0. A spring-loaded lever blocks an optical gate when filament is present. (b) Per-channel mechanical switches inside CANVAS, one detector board per channel, actuated by a spring-loaded bullet pin. Runout triggers a full flush macro and M2202 gcode_action_report=2505
2 sources
Tangle/snag detection (exact mechanism, if any — do not accept marketing wording at face value)Real hardware, not marketing. [tangling_switch_sensor] switch_pin=!toolhead:PA1. Physically: a rear-facing Hall-effect sensor at the top of the toolhead filament-detector board; a spring-loaded metal tab in the filament multiplexer extends when filament tension exceeds the spring force (the pneumatic-fitting hub compresses downward), triggering the sensor and raising a tangle error. Prevention side: CANVAS spool holders have a spring-loaded ratcheting rewind mechanism that takes up slack during unload
2 sources
Clog detection (mechanism, if any)Real: a filament-motion encoder. [z_filament_motion_sensor encoder_sensor] with extruder=extruder, detection_length=50 mm, switch_pin=PB4 (enable), motion_pin=PB3 (odometer), pause_on_runout=true. Mainboard pinout labels these S-DET PB4 = Filament motion detection and M-DET PB3 = Filament presence/absence. Elegoo lists Clog Detection: Yes for CC2, "–" for CC1 — this is a CC2-only feature
1 source
Filament change support (M600 / UI-driven)M600 is the pause g-code (machine_pause_gcode: M600). Multi-colour changes use Elegoo's own M6211 command (M6211 T[next_extruder] L[flush_length] M{old_filament_e_feedrate} N{new_filament_e_feedrate} Q[old_filament_temp] R[nozzle_temperature_range_high] S[new_filament_temp]), plus T[n] tool selects. On-screen load/unload is fully UI-driven
1 source
  • firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json (machine_pause_gcode, change_filament_gcode, machine_start_gcode M6211 A1 L200 …); firmware/CentauriCa · 2026-08 · local
Spool holder location & max spool size; drybox compatibilityFour spool holders bolt to the right-hand vertical extrusions (two tapped holes each), spools hang outside the enclosure, exposed to ambient air. Inner diameter 53–58 mm; larger spools need a printed adapter. Holders are spring-loaded with a ratcheting rewind. No drybox and no drying function anywhere in the product
1 source
Multi-material/multi-color capability or official upgrade pathNative: CANVAS, 4 colours, single nozzle. Firmware [canvas_dev] on /dev/ttyS1 @115200 with canvas_ctrl_pin=host:PD21; slicer support_mms: 1. The module is separable and also retrofits the CC1 ($55 CC1 version; $120–150 CC2/Centauri 2 version). No chaining/expansion beyond 4 channels
1 source
Cooling & filtration6 fields
Elegoo Centauri Carbon 2 — Cooling & Air Handling
Property and unitPublished value and conditionsSources and dates
Part-cooling fan(s): count, size, ducting designOne toolhead part-cooling fan: a custom 5020 radial blower with an integrated injection-moulded duct, P/N EFC-05D24D, Shenzhen Hua Xinrong, 0.50 A @ 24 V, 12 000 rpm, 4-pin (tach + PWM). Firmware [fan] pin=toolhead:PA8, tachometer_pin=toolhead:PB15, tachometer_ppr=2, tachometer_poll_interval=0.001
1 source
Auxiliary part-cooling fan?Yes — a custom 12032 radial aux fan on the left chamber panel, P/N EFC-12J24H, 0.8 A @ 24 V, 4-pin (tach + PWM), VHB-taped to the wall with an integrated moulded duct. Firmware [fan_generic fan1] pin=PD15, tachometer_pin=PD14; addressed as M106 P2. Slicer sets auxiliary_fan: 1
1 source
Heatbreak/hotend fan3010 axial fan, 3-pin with tachometer, 0.10 A @ 24 V, 12 000 rpm. Firmware [heater_fan heatbreak_cooling_fan] heater=extruder, pin=toolhead:PB13, tachometer_pin=toolhead:PB14, fan_speed=1. (always full when the extruder is on)
1 source
Chamber/exhaust fan & filtration (activated carbon / HEPA)Yes, and it is the CC2's headline thermal feature. A 12025 axial exhaust fan (P/N EFC12E24W, 0.5 A @ 24 V, 4-pin tach+PWM) works with a servo-driven shutter ("Smart Grille") that switches between exhaust-to-room and internal recirculation-through-filter. Filter media: HEPA + coconut-shell activated carbon ("designed to remove ultrafine particles, VOCs, and odors"), replaceable, $24.99 for 3. Firmware: [cavity_fan] pin=PD17, tachometer_pin=PD16 with a 5-step thermostat (T0=30, T1=38, T2=37, T3=40, T4=45 °C ↔ speeds A=1, B=0.7, C=0.3, D=1, E=0.3, F=0.4, G=0.5); [servo printer_servo] servo_pin=PB7, servo_det_pin=PB2, 1–2 ms pulse, maximum_servo_angle=180, initial_angle=90, M2801_open_angle=45 / M2801_open_time=10 ms, M2801_close_angle=45 / M2801_close_time=400 ms; a limit switch confirms grille position. Addressed as M106 P3
2 sources
Electronics bay coolingDedicated mainboard fan under stepper/heater-triggered control. [controller_fan board_cooling_fan] pin=PG15, tachometer_pin=PB9, stepper=stepper_x,stepper_y,stepper_z,extruder, heater=extruder,heater_bed, fan_speed=1.0, idle_speed=0.0, idle_timeout=30
1 source
Fan control (PWM? RPM feedback? firmware evidence)All five fans are PWM-controlled, and four of the five have tachometer feedback wired and polled in firmware: part-cooling (toolhead:PB15, ppr=2, poll 1 ms), heatbreak (toolhead:PB14, poll 1 ms), board fan (PB9), cavity/exhaust (PD16), aux (PD14). Elegoo markets a "Fan Self-Check" feature; the connectors are documented as "tach + 5 V PWM"
1 source
Supported materials6 fields
Elegoo Centauri Carbon 2 — Materials Support
Property and unitPublished value and conditionsSources and dates
Officially supported filament listElegoo's own two-tier list: "Ideal: PLA / PETG / ABS / ASA. Capable: TPU³ / PC / PA / PET / fiber-reinforced filaments⁴." The shipped slicer carries 30 CC2-specific filament profiles: PLA (Basic, PRO, +, Matte, Silk, Galaxy, Glow, Marble, Sparkle, Translucent, Wood, Rapid PLA+), PLA-CF, PETG (+ PRO, HF, Translucent, Rapid), PETG-CF, PETG-GF, PET-CF, ABS, ASA, ASA-CF, PC, PC-FR, PAHT-CF, TPU 95A, Rapid TPU 95A
1 source
PLA/PETG/TPU results & constraints (min TPU shore hardness)PLA and PETG: excellent and effortless — the unanimous reviewer verdict. Vendor profiles: PLA 210 °C / 21 mm³/s; PETG 250 °C / 11 mm³/s. TPU: works down to 95A, verified by Tom's ("The system worked great for every filament I printed with, including TPU down to 95A shore hardness"; a TPU Air Duster at 0.12 mm / 50 mm/s took "just over two hours … surface finish is a little rough, but the print is fully functional"). TPU constraint on the Combo: you must fit the Flexible Filament Adapter and physically disconnect the CANVAS cable — i.e. no multicolour TPU. Profiles: TPU 95A 3.2 mm³/s, PA 0.1
1 source
  • tomshardware.com (TPU 95A print, settings and result); firmware/ElegooSlicer_ECC2/Elegoo_TPU_95A_@ECC2.json, Elegoo_PLA_@ECC2.json, Elegoo_PETG_@ECC2.json; evidence/elego · 2026-08 · local
ABS/ASA feasibility (enclosure/chamber constraints, warping reports)Demonstrated on a US 110 V machine: a full standard-ABS print succeeded with no warping and no layer cracking. Aurora Tech printed standard ABS in a 17 C garage with the tall cover fitted; the job ran 3 h 41 min (about 18 % over the slicer estimate) and he reports "no corner warping", the part "sticks very well because the ambient temperature is stable", "no layer cracking", overhangs "printed nicely using standard ABS" and a good bottom surface, with the chamber holding ~45 C and peaking ~47 C (B6). Enclosure plus 110 C bed plus Smart Grille recirculation also give a "temperature hold mode where the vents shut off" for ASA (Notebookcheck). Vendor profiles: ABS 270 C / 18 mm3/s / fan capped at 40 %; ASA 270 C. Remaining caveats: (a) ASA is weaker than ABS here. Notebookcheck hit ASA adhesion problems, fixed them only by preheating the bed 10 minutes at 110 C, and photographed warping on an ASA test print; (b) there is no active chamber heating and no chamber preheat/wait, and an owner reports the CC2 starts printing before a requested 45-50 C chamber target is reached; (c) the ABS success is a single reviewer on a single part geometry. [RETRACTED after critique: an earlier draft claimed the bed mesh is probed at a fixed 60 C rather than at ABS bed temperature. It is not; the start g-code waits for print bed temperature (M190... A) before BED_MESH_CALIBRATE. See J6.]
4 sources
PC / Nylon / CF-filled feasibility (nozzle & extruder wear constraints)Demonstrated on real parts by Tom's Hardware, and this is the CC2's strongest surprise. PC: 7 utility knives in Prusa PC Blend, 0.2 mm, 100 mm/s, 6 h 06 min — "the prints were completed … and are perfect"; plus replacement spool holders in Polymax PC, 5 walls, 2 h 36 min. PA (nylon): Bambu Lab PAHT-CF RC part, 0.2 mm, 100 mm/s, 2 h 12 min — "Though this print can be a warping nightmare, it did really well". Tom's summary: "surprisingly strong with engineering materials, even without an actively heated chamber". Vendor profiles: PC 270 °C / bed 100 °C / 16 mm³/s / fan 35 % max; PAHT-CF 290 °C / bed 100 °C / 6 mm³/s / fan 30 %. Second independent confirmation (added 2026-08-19, reader-supplied): Aurora Tech, on a US 110 V unit, printed PA-CF successfully with the enclosure at ~41 C near the end of a 54-minute job, describing the surface as "as good as completely dried PACF" and the part as so stiff that "when I drop it on the table, it sounds like metal rather than plastic", despite the spool having been left open about two weeks; he also printed PPS-CF at the maximum nozzle temperature and reports it "worked very well" max, first 3 layers fan-off, flow ratio 0.96; PET-CF 270 °C / 8 mm³/s; PLA-CF 16 mm³/s with required_nozzle_HRC: 40. Abrasion is handled by the stock hardened-steel-tipped nozzle and SDK11 tool-steel hobbed gears; https://www.youtube.com/watch?v=xp-jQ7dljV0 at 21:20-22:10 and 26:20 (Aurora Tech; evidence/auroratech_cc2_combo_transcript.md)
1 source
Effective material ceiling given hotend+bed+chamber limits (reasoned, cite limits)Ceiling ≈ PC / PA(-CF) / PET-CF — i.e. everything the 350 °C nozzle and 110 °C bed can reach *that does not require a heated chamber*. Reasoning from the verified limits: nozzle firmware max 365 °C (claim 350) is enough for PC (270), PAHT-CF (290) and PET-CF (270) with margin, and would nominally reach PPS/PPA-class temperatures; bed firmware max 125 °C (claim 110, ot_bed=130) covers PC/PA bed requirements (vendor profiles use 100 °C); the binding constraint is the chamber — passive only, thermostat setpoints imply 30–45 °C, ot_box=100, no heater, no preheat-and-wait. That rules out reliable large ABS/ASA/PC-CF/PPS parts and warp-sensitive geometry, which is precisely where 50–60 °C active chambers earn their keep. Practically: excellent to PETG/PLA-CF; very good for small-to-medium PC and PA-CF; marginal for large ABS/ASA; not a PPS/PEEK machine despite the nozzle number — notwithstanding Elegoo's own non-Combo comparison table, which lists "Capable: PPS/PPS-CF/PPS-GF, PPA-CF/PPA-GF" (see R1); that claim is rejected here on the chamber argument and no reviewer has printed PPS on either Centauri
1 source
Filament profile ecosystem in the recommended slicerStrong and CC2-specific. ElegooSlicer (an OrcaSlicer/BambuStudio fork, distributed on the included USB and on GitHub) ships 30 filament, 21 process (0.08–0.48 mm layers across 0.2/0.4/0.6/0.8 nozzles) and 5 machine profiles named specifically for the CC2 (@ECC2 / @Elegoo CC2). Per-material pressure advance, flow caps, fan curves, plate temperatures and exhaust-fan logic are all pre-set. OrcaSlicer and Cura are also listed as supported. RFID auto-selects the profile for Elegoo-branded spools
1 source

Software & workflow

Firmware, controls, connections and calibration.

Electronics & power6 fields
Elegoo Centauri Carbon 2 — Electronics & Power
Property and unitPublished value and conditionsSources and dates
Mainboard model & MCU (exact chip markings if teardowns exist)Elegoo "self-developed" mainboard, part number EF-M5-V1.0.073. Main SoC: Allwinner R528-S3, which *also* runs the motion MCU firmware on its internal HiFi4 DSP core — the shipping config's [mcu] is serial=/dev/rpmsg_ctrl-dsp_rproc@0 with restart_method=remoterpc, i.e. rpmsg to an on-die DSP, not an external microcontroller. Two auxiliary MCUs: toolhead = GD32 (firmware build target toolhead_gd32, version string toolhead_gd32-00.00.00.03-app), bed = STM32F402RCT6. CANVAS mainboard = GD32F303RCT6
3 sources
Host computer (SoC, RAM, storage) if firmware runs on embedded LinuxAllwinner R528-S3, 32-bit ARM (arch/arm, sun8iw20p1 platform, SMP, VFP+NEON), 128 MB in-package DRAM, 8 GB eMMC. Linux 5.4.61 (Allwinner BSP). Rootfs uses opkg (OpenWrt-style ipk packaging) with app data under /opt (/opt/inst, /opt/usr/cfg, /opt/usr/gcode)
2 sources
PSU (brand, wattage, rail voltage)Partially known. Logic/motor rail is 24 V DC (Elegoo spec "Power Output: 24 V"; mainboard input is a 2-pin barrier terminal marked +24 V/GND with 9.6 mm pitch). The bed does not use the 24 V rail at all — it is switched mains via an SSR board. PSU brand, model and 24 V wattage are
2 sources
Input voltage (110/220 V, switch or auto)AC 100–240 V, 50/60 Hz — auto-ranging, no voltage switch documented
1 source
  • evidence/elegoo_official_product_page.md ("Power Input: 100–240 VAC 50/60 Hz"); docs/CC2_manual_EN.pdf p.01; techradar.com ("Input Power: AC 100–240 V, 50/60 Hz"); notebo · 2026-08 · local
Total rated power draw1100 W @ 220 V; 350 W @ 110 V (rated, not measured)
1 source
Expansion headers / spare ports (fans, steppers, GPIO, USB host)Unusually well documented. Mainboard carries: 2×4 pin header = DSP UART + UART0 serial consoles (5 V/TX/RX/GND each); a FEL 4-pin header (GND/DP/DM/5 V) for Allwinner recovery; a front USB-A host port; a camera 4-pin USB; the CANVAS "AMS port" (2×2 Micro-Fit 3.0, RS485-A/B + 24 V + GND); Bed MCU RS-232 (5-pin, 5 V switched for reset); JST-XHB connectors for chamber temp (PH0), light (PB5), FAN-1 (PD14/PD15), BOX-F (PD16/PD17), exhaust servo (PB2/PB7 + limit switch), Z-endstop (PB10), bed heater PWM (PG13), bed temp (PB14), mainboard fan; and an onboard piezo buzzer (PB12). Toolhead link is a USB-C connector carrying 24 V on Vbus
2 sources
Firmware & openness12 fields
Elegoo Centauri Carbon 2 — Firmware & Openness
Property and unitPublished value and conditionsSources and dates
Firmware base & current version (Klipper fork / Marlin fork / proprietary)"ELEGOO OS" = a closed C++ re-implementation of Klipper. The released source tree is unmistakably Klipper: src/ + lib/ with the full Klipper MCU HAL set (stm32, rp2040, atsam, avr, ar100, linux, simulator, lib/kconfiglib), and elegoo/ containing printer.cpp, configfile.cpp, chelper/, kinematics/, extras/ (215 files) — the Klippy host transpiled from Python to C++. Config syntax is Klipper's ([stepper_x], [tmc2209 …], [input_shaper], [bed_mesh], rotation_distance, pressure_advance). Slicer declares gcode_flavor: klipper. Community binary analysis pins the base at Klipper commit 28f60f7e, 616 commits past v0.9.1 — ⚠ this specific commit identification is a CC1 finding: freethecode.lol is a Centauri Carbon (CC1) campaign page and mentions the CC2 zero times (verified by full-text search, 2026-08-21). It is quoted here as the platform lineage, not as a CC2-specific binary analysis; nobody has published a commit-level diff of a CC2 binary. Latest OTA catalogued: v02.00.02.00 (2026-05-28); the published source corresponds to ~v01.01.13.10 / release 2602101855
3 sources
Source code availability (repos, GPL compliance status, last update date)A GPL-3.0 source repo exists but is stale and incomplete, and compliance is actively disputed. elegooofficial/CentauriCarbon2 was created 2026-03-11, last pushed 2026-03-12, and has never been updated — while at least three OTA firmware releases have shipped since (v01.03.02.36, v01.03.02.51, v02.00.02.00). Open issue #8 "Out of compliance with GPLv3" (2026-06-02) is still open, with 8 comments escalating through 2026-08-14 — including an @-mention of @realrossmanngroup on 2026-08-13 and, on 2026-08-14, the first reply from an ELEGOO account (anjis-elegoo), which quotes the complaint back but does not commit to a source-release date. Re-verified 2026-08-21: still open, no new source push. Open issue #1 shows the release cannot even be built: mcu_build.sh -b toolhead_gd32 fails with can't found configfile: CentauriCarbon2/mcu_config/toolhead_gd32.config — the directory is missing from the repo. The same "Out of compliance with GPLv3" issue is open on the CC1 repo (#26). The freethecode.lol campaign is still live and still marked "Still ongoing" (issue raised 2025-08-07)
7 sources
Root/SSH access: possible? official or community methodAnd this is the answer that changed in mid-2026. An SSH daemon was present in CC2 firmware up to and including v01.03.02.51 (2026-05-09). Firmware v02.00.02.00, released 2026-05-28, "removes the SSH process entirely" and simultaneously blocks downgrades to any earlier version. OpenCentauri publishes a repacked v01.03.02.51 specifically to defeat the downgrade block ("Use at your own risk"). No CC2 community firmware currently restores SSH: OpenCentauri Patched Firmware is explicitly "Not currently" compatible with the CC2 (it is built on CC1 1.1.40), and COSMOS's public FAQ still says CC2 support is a "Maybe". Hardware-level root remains available via Allwinner FEL mode + UART (documented for the CC2 specifically, eMMC on mmc dev 1)
5 sources
Moonraker / Mainsail / Fluidd or other open API available (stock vs after mods)?None of them, stock. The host is a single C++ binary with a proprietary IPC socket (-a /tmp/elegoo_uds) and an embedded Mongoose 7.12 web server. What *does* exist: a built-in WebUI, an MJPEG camera stream on port 3031 (:3031/video), an Elegoo SDCP protocol used by ElegooSlicer/the Matrix app, an MQTT transport (the non-shipping config even hard-codes broker_address=tcp://broker.emqx.io:1883), and OctoEverywhere / Home Assistant integrations built on the SDCP API. OpenCentauri has reverse-engineered and documented the API and MQTT surface
4 sources
Notable community firmware/mods (names, repos, what they add)For the CC2 specifically: none usable today. The ecosystem is real but CC1-first. (a) OpenCentauri Patched Firmware (repo OpenCentauri/cc-fw-tools) — adds SSH root, opkg/entware, boot logo, homing-position change, mid-print WebUI edits, USB-Ethernet support, M8212/M8213 chamber-light g-codes and TEMPERATURE_WAIT SENSOR=box; CC1 only. (b) COSMOS (OpenCentauri/cosmos, 231 stars) — full Klipper/Kalico on stock hardware via Yocto; "Add support for CC2" was merged 2026-07-01 (PR #225, author's own note: *"Untested, will still need some work probably"*), release 26.07.0 shipped 2026-07-29, yet the public FAQ still says CC2 is a "Maybe". (c) OpenCentauri/cad — full CC2 CAD models and FDM-optimised replacement parts. (d) Elegoo-Tag-Library — dumps of Elegoo RFID tags. (e) An OpenCentauri drop-in replacement mainboard project is in progress
8 sources
Firmware update mechanism (OTA/USB) & update cadence historyOTA over Wi-Fi, or offline via USB. Packages are.zip.sig — a signed/encrypted wrapper containing *.swu.sig (the swupdate payload) and *.json.sig (metadata). The device uses an A/B partition scheme managed by swupdate: the update is written to the inactive slot which becomes primary on next boot. USB path: drop the.zip.sig on a FAT32 stick → Settings → Check for Updates → Offline Update. Cadence: 7 releases in ~7 months (2025-10-31, 2025-11-17, 2025-12-11, 2026-01-09, 2026-04-09, 2026-05-09, 2026-05-28)
1 source
G-code flavor & notable custom G-codes/macrosKlipper flavour (gcode_flavor: klipper) plus a substantial proprietary extension set: G180 S<n> — a numbered macro dispatch table (S1–S10 calibration steps, S800 "move to waste box", S810–S813 flush/cutter steps, S900-series mesh profile logic); M6211 — CANVAS tool change with flush length, feedrates and temperatures; M2202 gcode_action_report=<code> — UI progress reporting; M2801 — exhaust-grille servo open/close; M190 S<t> A — bed heat with an extra A flag; SET_CONDITION J/A/B … SET_CONDITION_END — a conditional-execution primitive not present in mainline Klipper; plus standard BED_MESH_CALIBRATE, SAVE_CONFIG, SET_PRINT_STATS_INFO, PROBE, SET_KINEMATIC_POSITION, TEMPERATURE_WAIT. OpenCentauri maintains a custom-g-code reference
1 source
Root shell: exact method (official setting toggle / vendor-documented / community exploit / none). Is root persistent across OTA firmware updates, or wiped each update?Three tiers, none of them a supported toggle. (1) No official/vendor-documented root method exists — there is no developer-mode switch in the UI and no Elegoo documentation of shell access. (2) Firmware-resident SSH existed up to v01.03.02.51 (2026-05-09) and was deleted in v02.00.02.00 (2026-05-28); OpenCentauri describe the change as removing "the SSH process entirely, making it more difficult to perform research on a live device", and the same release blocks downgrades. Staying on ≤v01.03.02.51, or flashing OpenCentauri's repacked v01.03.02.51, is therefore the only firmware-level route — and I could not find published CC2 root credentials (the widely-quoted root / OpenCentauri pair belongs to OpenCentauri's CC1 patched firmware, which does not run on a CC2). (3) Hardware-level root is confirmed available: Allwinner FEL mode (short R53 while resetting, or efex from U-Boot) plus the on-board UART0/DSP serial headers, with CC2-specific artefacts published (u-boot-sunxi-with-spl-cc2.bin, backup_cc2.scr/restore_cc2.scr, eMMC on mmc dev 1) — this gives full eMMC read/write, i.e. offline root. Persistence: an OTA update replaces the entire rootfs slot (A/B swupdate), so any rootfs modification is wiped by the next official update.
6 sources
Userland: OS base (Debian/Ubuntu/Buildroot/OpenWrt/busybox-only), package manager present (`apt`/`opkg`/`ipkg`/none), CPU architecture (armv7 / aarch64 / mips / x86), and free space on the writable partition. A busybox/no-apt rootfs with a few MB free is a red flag — say so.OS base: a Buildroot/OpenWrt-style embedded Linux, not Debian/Ubuntu. Evidence: opkg is the package manager (Elegoo's own build docs: "开发板上opkg常用指令 … opkg install xxx.ipk", -o /opt/packages/), packaging is ipk with control.tar.gz/data.tar.gz and debian-binary = 2.0, init is /etc/init.d-style shell scripts, and app state lives under /opt (/opt/inst, /opt/usr/cfg, /opt/usr/gcode). Kernel: Linux 5.4.61 (Allwinner BSP). Architecture: 32-bit ARM (armv7-class), arch/arm, platform sun8iw20p1, SMP, VFP+NEON — *not* aarch64. RAM: 128 MB total, in-package. Storage: 8 GB eMMC. Free space on the writable partition: for the CC2 — the CC1-derived partition map gives rootfsA/B ≈ 134 MB each, rootfs_data ≈ 134 MB, UDISK ≈ 7.06 GB, but I did not find a CC2 partition dump
3 sources
Native Tailscale feasibility: can `tailscaled` run persistently on-device? Cite concrete evidence — community guides, wiki entries, forum posts, or repos of people actually doing it on THIS model. Note whether it needs `/dev/net/tun` (kernel TUN support) and whether the kernel provides it, plus whether userspace-networking mode is the only option. Distinguish "someone did it" from "should theoretically work."Not demonstrated by anyone, and structurally hostile. (a) No evidence any person has run tailscaled on a Centauri Carbon of either generation — I searched OpenCentauri's docs, its 4 relevant GitHub orgs/repos' issues, freethecode.lol, and the open web for "Tailscale + Centauri Carbon"; the only 3D-printer Tailscale guidance found is generic (Voron/Raspberry-Pi subnet-router patterns). (b) /dev/net/tun is very likely absent: CONFIG_TUN is not set in either Allwinner R528/T113 defconfig in OpenCentauri's fork of the vendor kernel (sun8iw20p1smp_defconfig, sun8iw20p1smp_t113_auto_defconfig); the only tunnel symbol present is CONFIG_IPV6_TUNNEL=y, and no CONFIG_NETFILTER* symbols are set at all. Without TUN, tailscaled must run --tun=userspace-networking, which rules out subnet-router and exit-node roles and costs throughput. (c) 128 MB RAM shared with the motion controller is a hard practical ceiling — tailscaled's resident set is a large fraction of that. (d) On current firmware there is no shell to install it from (N8), no apt, and an A/B rootfs that erases changes on update (N12)
4 sources
SSH daemon: present by default or must be installed/enabled? Default credentials if documented. Does an OTA update disable it or reset the password?Present by default up to v01.03.02.51; absent from v02.00.02.00 onward. OpenCentauri's wording is that v02.00.02.00 "removes the SSH process entirely", i.e. the daemon shipped enabled (or at least resident) in earlier builds. Default credentials for stock CC2 firmware: — I found none published. The frequently-cited root / OpenCentauri pair is explicitly the credential set for OpenCentauri's patched CC1 firmware, which does not support the CC2. An OTA update to v02.00.02.00 removes SSH and blocks rollback, so an OTA can and does permanently disable it
3 sources
Filesystem persistence: is the rootfs read-only or overlay-mounted? Do user-installed binaries and init/systemd services survive reboot and survive firmware updates? Is there a supported place to persist customizations?A/B rootfs managed by swupdate; user-installed software does not survive official updates. Each update is written to the *inactive* rootfs slot, which then becomes primary — so anything added to the live rootfs is left behind on the now-inactive slot. There is a separate rootfs_data partition (overlay/persistent data) and a large UDISK partition; Elegoo's own supported persistence locations are /opt/usr/cfg (user config, incl. autosave.cfg written by SAVE_CONFIG) and /opt/usr/gcode (job storage), with defaults in /opt/inst. run_printer.sh even contains a commented-out "copy default config to user config" path, showing Elegoo intended /opt/usr/cfg/user_printer_dsp.cfg as a user-override slot — but it is commented out in the shipped script, so the binary is launched against /opt/inst/printer_dsp.cfg and user config overrides are not active
2 sources
Connectivity, display & camera7 fields
Elegoo Centauri Carbon 2 — Connectivity, UI & Camera
Property and unitPublished value and conditionsSources and dates
Display: size, type, touch?5.0-inch colour capacitive touchscreen, front-mounted, connected by FFC ribbon (RGB888 + touch). Controller is a Sitronix ST77922 panel with an ST1633i I²C capacitive touch controller
4 sources
On-device UI capabilities (model preview, per-object controls, etc.)Model preview and file browse (local + USB), one-click calibration suite, per-spool CANVAS load/unload with manual filament type/colour entry, RFID tap-to-read, camera live view + timelapse, chamber light toggle, Z-offset adjust, network/IP display, offline firmware update, print history. v01.03.02.51 added "the ability to exclude and reset regions containing foreign objects" (a foreign-object exclusion feature). Per-object cancellation ("skip object") is NOT supported — it is an open feature request
2 sources
WiFi (band/standard), Ethernet, USB ports, SD/flash-driveWi-Fi (2.4 GHz-class AIC8800D80 SDIO module) + USB flash drive only. NO Ethernet port. No SD card slot. Hardware: a front USB-A host port and a separate 4-pin camera USB header on the mainboard
2 sources
Cloud service & phone app; LAN-only mode possible?Cloud: Nexprint (Elegoo's model hub, integrated into ElegooSlicer's home screen, in beta) and the ELEGOO Matrix mobile app (binding added in v01.03.01.89, app ≥ v1.0.11). LAN-only mode exists and works: connect to Wi-Fi, note the displayed IP, add the printer in ElegooSlicer's Device tab on the same subnet. However, LAN-exclusive mode is mutually exclusive with useful features
2 sources
Camera: built-in / optional? Resolution & framerate (verify — marketing often omits real fps)Built-in and standard. 1280 × 720 (720p), 16:9, max 30 Hz, fixed focus (glued barrel), auto exposure, FOV ~90–110°, USB 2.0, 30 × 30 mm PCB, 500 mA @ 5 V, 4-pin JST-ZH 1.5 mm connector, enumerates as "Integrated Camera". Stream is MJPEG on port 3031 (:3031/video)
2 sources
Camera features: live monitor, timelapse, AI failure detection (spaghetti detection) — how real is it?All three present. Live view in the WebUI/Matrix app/ElegooSlicer; timelapse with video export (a v02.00.02.00 changelog entry fixes "mismatch between timelapse video filenames and print file filenames during export"); AI spaghetti/failure detection with notifications. Real-world reliability is mediocre
2 sources
LED chamber/toolhead lightingChamber LED present and firmware-controlled: [output_pin led_pin] pin=PB5, value=0; mainboard "Light" connector is a JST-XHB-3P with +24 V (max 1 A), GND and signal on PB5. Toolhead has an LED PWM pin that is not connected
2 sources
Calibration & everyday use7 fields
Elegoo Centauri Carbon 2 — Calibration Routines & QoL
Property and unitPublished value and conditionsSources and dates
One-touch auto-calibration suite: what exactly does it calibrate?Elegoo "Full-Auto Calibration". Verified contents: (a) 121-point bed mesh, (b) Auto Z-offset via load-cell nozzle touch, (c) input-shaping / vibration compensation, (d) an automated nozzle-clean cycle that precedes probing, (e) nozzle-temperature Z-compensation table, (f) sensor self-checks (fan, exhaust grille, cutter status, CANVAS). Notebookcheck timed the whole thing at 26–27 minutes; CreativeBloq at "around ten minutes" and notes "it gets somewhat noisy"
2 sources
PID autotune exposed to user?Yes — there is a dedicated [pid_calibrate] section with its own preparation macro (pid_calibrate_gcode=G90\nG28 X Y TRY\nM204 S10000\nG180 S800\nM400\n, i.e. home and park at the waste station before tuning). Factory PID values are shipped for both heaters: extruder pid_kp=25.989 / ki=3.397 / kd=49.705; bed pid_Kp=46.504839 / Ki=1.453276 / Kd=372.038710
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L444–445 ([pid_calibrate]), L128–131 (extruder PID), L158–161 (bed PID) · 2026-08 · local
Flow-rate calibration toolingAs an on-printer feature. No flow-calibration macro exists in [g180] or elsewhere in the config. The practical path is ElegooSlicer's inherited OrcaSlicer calibration suite (flow ratio, PA, temperature towers), since ElegooSlicer is an OrcaSlicer/BambuStudio fork. Vendor per-material flow ratios are shipped (e.g. PAHT-CF filament_flow_ratio: 0.96)
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg (searched: no flow-calibration macro); firmware/ElegooSlicer_ECC2/Elegoo_PAHT-CF_@ECC2.json (filament_flow_ratio); toms · 2026-08 · local
Belt tension check/indicatorsNone. No belt-tension routine, indicator, or resonance-based tension readout in firmware; no tensioning procedure in the manual. Hardware provides spring-loaded automatic tensioners instead (BOM X/Y轴同步带张紧弹簧 = XY belt tensioning springs, XY轴皮带张紧惰轮支架 = tensioner idler brackets)
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg (no tension macro); docs/CC2_manual_EN.pdf (no tensioning section); docs/CC2_BOM.pdf (tensioning springs and idler brac · 2026-08 · local
First-layer inspection featuresYes — scan_first_layer: 1 is set in the machine profile, i.e. camera-based first-layer inspection is enabled. Separately, an elaborate automated nozzle-clean-and-verify routine runs before probing (clean_nozzle_gcode / [g180] S1–S2: heat to 200 °C, wipe on the silicone wiper, cool to 140 °C, run a Z-touch pattern, re-wipe)
1 source
  • firmware/ElegooSlicer_ECC2/Elegoo_Centauri_Carbon_2_0.4_nozzle.json (scan_first_layer: 1); firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L644+ ([g180] S1, S2, S911) · 2026-08 · local
Skew calibrationNone. No [skew_correction] section in either config, no skew routine in [g180], nothing in the manual or any review
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg and debugfile_printer.cfg (full section lists — no skew section); docs/CC2_manual_EN.pdf · 2026-08 · local
Assembly level & time from unbox to first printSemi-assembled; ~15–20 minutes of assembly, ~45 minutes unbox-to-first-print including calibration. Steps: remove 3 bed transport screws, clip on the touchscreen (ribbon cable), screw on the nozzle-wiper assembly and waste chute (4 screws, from inside, at the rear), bolt the CANVAS bracket + module (3 screws), plug the CANVAS power/signal cable, route 4 Bowden tubes to the toolhead hub (2 screws), bolt 4 spool-holder brackets (2 screws each), fit the tophat + CANVAS extension
2 sources
Interruption recovery3 fields
Elegoo Centauri Carbon 2 — Power-Loss & Interruption Recovery
Property and unitPublished value and conditionsSources and dates
Power-loss recovery: mechanism (how is state saved? capacitor / eeprom / file on shutdown?)Real hardware-sensed power-loss detection with an energy-reserve circuit, not a periodic-file-write hack. [por] (power-outage recovery): sensor_pin=host:PG9 (mains-loss sense), ctrl_pin=host:PG11, charge_pin=host:PD20, charge_time=30.0 (a reservoir/supercap charge line), rest_time=0.01 (10 ms poll), and an explicit shutdown_pins list that de-energises everything in a defined order on detection: host:PD22 (LCD PWM), host:PE17 (nozzle power), host:PB11 (bed power), PG13 (bed control), !PF6/!PE15/!PE12 (X/Y/Z enables), PB5 (light), PD17/PD19 (box fan), PD15 (fan1), PG15 (mainboard fan). Enabled at the print layer via [virtual_sdcard] power_outage_enable=1
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L607–630 ([por] incl. charge_pin, charge_time, shutdown_pins), L223 (power_outage_enable=1) · 2026-08 · local
Verified working? Third-party test results; resume seam qualityConfirmed working by an independent reviewer, but with documented bugs that took two firmware releases to fix. Notebookcheck: "the auto-filament refill, filament runout sensor, hotend cover sensor, and power loss recovery … worked as they should." Firmware history shows real defects: v01.03.02.51 (2026-05-09) fixed "print history showed a total print time of 0 after resuming printing following a power outage", and v02.00.02.00 (2026-05-28) fixed "resume printing failed after power loss when using an external spool under certain conditions". v01.03.02.36 "Optimized print recovery quality after pause/resume". No reviewer photographed or measured the resume seam
2 sources
Filament-runout pause/resume behaviorAutomatic and elaborate. On runout the printer runs a full macro: M400 → G28 X Y TRY → M204 S10000 → park at the waste station (G180 S800) → part fan to 25 % → purge E10 F300 + E5 F150 → fan to 88 % (M106 S225) → 5 s dwell → three-stage flush/wipe (G180 S811 → S810 → S800) → fans off → report M2202 gcode_action_report=2505 to the UI. Governed by pause_on_runout=true, event_delay=10.0, pause_delay=1.0. In CANVAS mode, auto-refill from a duplicate spool takes over instead of pausing. Pause/resume additionally has a scripted purge-and-prime on resume
2 sources

Safety, support & testing

Published safeguards, support terms and independent evidence.

Safety5 fields
Elegoo Centauri Carbon 2 — Safety
Property and unitPublished value and conditionsSources and dates
Thermal runaway protection (verify in firmware source if at all possible)Present and firmware-verified for both heaters. [verify_heater heater_bed]: max_error=60, check_gain_time=60, hysteresis=5, heating_gain=2. [verify_heater extruder]: max_error=30, check_gain_time=20, hysteresis=5, heating_gain=2. These are Klipper's standard runaway semantics (heater must gain ≥2 °C per check_gain_time while heating; cumulative error beyond max_error faults the printer). Backed by independent hardware cutoffs: ot_bed=130 and ot_box=100, and a thermal fuse in the mains bed circuit
1 source
Certifications (UL/ETL/CE/FCC — verify, don't assume)I could not verify a single certification for the CC2. The printed manual contains no FCC/CE/UKCA/RoHS/WEEE statement and no declaration of conformity; the Elegoo product page lists none; the packaging/nameplate was not available to me
1 source
  • docs/CC2_manual_EN.pdf (searched full extracted text for "FCC", "CE", "UKCA", "RoHS", "WEEE", "UL", "ETL", "certif", "complian", "declaration" — zero hits); evidence/eleg · 2026-08 · local
Hardware safety: fusing, strain reliefs, belt covers, pinch pointsPartly documented. Bed circuit: mains AC with ground present and a thermal fuse, switched by an SSR board. OpenCentauri's own warning is unambiguous: "Danger: The Bed runs on mains AC current at 110/220 V. These voltages may be lethal. Exercise extreme caution around the bed and SSR board." Enclosure with glass door and tophat covers all belts and the gantry. Manual safety notes cover burns, pinch hazards, grounding, dust maintenance and adult supervision of children. Mainboard 24 V input is a 2-pin barrier terminal described as "very flimsy — do not overtighten." Toolhead USB-C carries 24 V on Vbus (fries anything else plugged in). No fuse ratings or strain-relief details are published
2 sources
Firmware safety limits (max temp cutoffs — exact values)Nozzle max_temp=365 / min_temp=-100; bed max_temp=125 / min_temp=-100, ot_bed=130; chamber max_temp=110 / min_temp=-100, ot_box=100. Idle timeouts: extruder 300 s (extruder_timeout_gcode=M104 S0), bed 600 s (range 60–259 200 s). Probe force limits: trigger_force=200 g, safety_limit=1200 g. min_extrude_temp=-100 — cold-extrusion protection is DISABLED, as is Klipper's extrusion cross-section check (max_extrude_cross_section=9999999)
1 source
  • firmware/CentauriCarbon2/debugfile_printer_dsp.cfg L133–134 (extruder), L163–165 (bed), L256–258 (chamber), L437–442 ([idle_timeout]), L536–537 (probe force), L141–142 (m · 2026-08 · local
Known safety incidents / recalls / fire reports (with sources)None found for the CC2. No recall, fire report or safety incident surfaced in any of the eight reviews, the Elegoo GitHub issue trackers (CC2 repo: 11 issues, none safety-related), the OpenCentauri issue trackers, freethecode.lol, or general web search
3 sources
Noise & power consumption3 fields
Elegoo Centauri Carbon 2 — Noise & Power Consumption
Property and unitPublished value and conditionsSources and dates
Noise: claimed dB and independently measured dB (distance/method)Claimed ≤45 dB (footnoted "Data from Elegoo Lab. Tested with Elegoo PLA filament in Silent Mode with the auxiliary part cooling fan turned off"). Independently measured by Notebookcheck: average 44–47 dB, peaks to 54 dB. Distance and instrument not stated
1 source
Silent/quiet mode available?Yes — Elegoo's own noise footnote references "Silent Mode". Firmware side, the mainboard fan idles at zero (idle_speed=0.0, idle_timeout=30) and the cavity fan is thermostatically staged, both of which reduce idle noise. No dedicated silent-mode section appears in the released config, so the mode is implemented above the motion layer (UI/host)
1 source
  • evidence/elegoo_official_product_page.md (footnote 2: "Tested with Elegoo PLA filament in Silent Mode with the auxiliary part cooling fan turned off"); firmware/CentauriC · 2026-08 · local
Power draw: idle / PLA print / high-temp print (measured preferred)No measured CC2 power figures exist. Rated only: 1100 W @ 220 V; 350 W @ 110 V, 24 V logic rail, ~1000 W of that being the mains bed
1 source
  • Searched all eight CC2 reviews plus targeted web search for CC2 wattage/power-meter tests; evidence/elegoo_official_product_page.md; docs/CC2_manual_EN.pdf p.01 · 2026-08 · local
Support & ecosystem6 fields
Elegoo Centauri Carbon 2 — Reliability, Support & Ecosystem
Property and unitPublished value and conditionsSources and dates
Warranty length & support reputation (evidence, not vibes)12 months standard on the printer, with a component-tiered warranty table: consumables carry no warranty (PEI platform plate, build plate, nozzle kit, PTFE tube, heater, thermistor, hot end, packaging, manual, tools); CANVAS parts have a limited warranty covering the feeder-module motherboard, 4-pin cable, five-way module, spool holder (excluding support claws), feeder motor, communication board, gearbox/cutter assembly, fan and housing, with "damage caused by human error … not covered". A 2-year extended warranty is sold for the Centauri series (e.g. via MatterHackers). Support reputation: mixed, evidence-based
2 sources
Common failure points (from long-term reviews/forums, with sources)Six documented so far, all from the first ~7 months. (1) Spool holders — Tom's physically broke one and "tore the rubber grips off" with a third-party cardboard spool; TechRadar had Elegoo plastic reels slip off the holder; OpenCentauri documents a ratchet clicking noise needing grease. (2) Toolhead filament-hub collets are effectively unremovable — Tom's could not extract the PTFE tubes even with the supplied tool and help, so clearing an extruder clog means unscrewing the whole hub. (3) 250 °C load routine cannot purge high-temp residue (firmware-confirmed). (4) Software/connectivity flakiness — Notebookcheck: random refusal to connect to ElegooSlicer on Windows, monitoring-window errors, 17 %-optimistic time estimates; GitHub issue #9 LAN-mode exclusivity. (5) False positives on clog/spaghetti and tangle detection (Notebookcheck; firmware v01.01.16.40 changelog). (6) Purge waste volume — 3DwithUS: on a 4-colour Gengar print "the purge waste was considerable though, almost the same amount as the model itself"
4 sources
Spare parts availability & pricing (nozzle, hotend, plate, board, belts)Good and CC2-specific, ordered direct from Elegoo US. Verified live prices (2026-08-19): 0.4 mm brass/hardened-steel nozzle kit $22.99; multi-size (0.2/0.4/0.6/0.8) nozzle kit $24.99; dual-sided build plate $22.99 (3-pack $39.99); ceramic heater (CC2) $9.99; thermistor (CC2) $9.99; hotend silicone sock ×3 (CC2) $9.99; composite HEPA+carbon filter ×3 (CC2) $24.99; auxiliary part-cooling fan (CC2) $28.99; 3010 axial fan $12.99; dual extruder gears $14.99; strain gauge $12.99; spool holder (CC2 Combo) $21.99, set of 4 $45.99; spool claws ×3 $10.99; anti-vibration feet ×4 (CC2) $11.99; CANVAS module $150 (with cover) / $120 (without); Accessory Kit for Centauri Series $59.99
1 source
Community size (subreddit size, wiki quality, Discord/Facebook groups)Substantial and unusually technical for a 7-month-old machine. OpenCentauri Discord ~5,800 members; ELEGOO Official Discord ~32,700 members. OpenCentauri org: 26 public repos, cosmos at 231 stars with active development through 2026-08-21, plus a full MkDocs wiki with dedicated CC2 hardware pages (mainboard, toolhead, bed, camera, screen, misc, CANVAS, docs) and a CC2 firmware archive. Elegoo maintains official GitHub repos with public issue trackers for both CC1 (25 issues, numbering to #28 because three numbers are PRs) and CC2 (11 issues, all open). Elegoo also runs official Reddit and Facebook presences
6 sources
Mod/upgrade ecosystemActive but CC1-weighted. CC2-applicable today: full CAD models and FDM-optimised replacement parts (OpenCentauri/cad, 465 MB repo, pushed 2026-07-06, includes a CC2 model per PR #13); printable mods, light mods and camera replacement guides; Elegoo RFID tag dumps (Elegoo-Tag-Library); CANVAS retrofit onto CC1; printable spool holders (Tom's used a Voron-derived design after breaking his). Not yet CC2-ready: OpenCentauri patched firmware (CC1 only), COSMOS (CC2 support merged 2026-07-01 but "untested" and not endorsed in the FAQ), the OpenCentauri replacement mainboard project
5 sources
Long-term (6–12 month) reliability reportsNone exist yet. The CC2 shipped 2026-01-26, so the oldest retail units are ~7 months old at the access date. The longest documented exposure I found is 3DTechValley's "3 weeks of hands-on testing with a retail unit"; every other review is a launch-window unit tested over days. No 6-month or 12-month follow-up, durability teardown, or MTBF data exists
2 sources
Independent measurements5 fields
Elegoo Centauri Carbon 2 — Independently Measured Performance
Property and unitPublished value and conditionsSources and dates
Dimensional accuracy measurements (third-party, with numbers)TechRadar, calipers on a 5-step accuracy target — the only real numbers that exist. Target 25 mm → X 24.93 (−0.07), Y 24.85 (−0.15); 20 mm → X 19.97 (−0.03), Y 20.01 (+0.01); 15 mm → X 14.94 (−0.06), Y 14.95 (−0.05); 10 mm → X 9.96 (−0.04), Y 9.88 (−0.12); 5 mm → X 4.99 (−0.01), Y 5.01 (+0.01). X mean error 0.042 mm; Y mean error 0.068 mm; combined mean 0.055 mm. Scored 5/5. Elegoo's claim is ±0.1 mm
1 source
Surface quality / VFA / ringing assessmentsThe only structured test scores it poorly, while every impressionistic review praises it. TechRadar's sub-scores: Fine flow control 5/5, Fine negative features 5/5, Bridging 5/5, Overhangs 4/5, but XY resonance 2.5/5 and Z-axis alignment 2.5/5. Against that: Tom's "no color bleeding using the default purge values"; CreativeBloq "layer lines are barely visible, corners and edges are sharp" and "very few of those annoying wispy bits of filament"; 3DwithUS "All lettering came out sharp and clearly defined" at 0.08 mm; TechRadar's own prose "The multifilament printing quality is exceptionally good, with clean lines"
2 sources
Benchmark print times (source + settings)Tom's Hardware is the only outlet that published settings alongside times: 7× Baby Chameleon (multicolour PLA, 0.2 mm, 160 mm/s avg) 4 h 55 min; Slinky Fidget (PETG, 4 walls, 50 % infill, 0.2 mm, 150 mm/s avg) 4 h 20 min; TPU Air Duster (TPU 95A, 0.12 mm, 50 mm/s) ~2 h; RC Death Racer part (Bambu PAHT-CF, 0.2 mm, 100 mm/s) 2 h 12 min; 7× Utility Knife (Prusa PC Blend, 0.2 mm, 100 mm/s) 6 h 06 min; Voron spool holders (Polymax PC, 5 walls, 0.2 mm, 100 mm/s) 2 h 36 min. 3DwithUS: 4-colour Gengar, 0.14 mm, 4 colours, ~5 in tall, ~14 h; HueForge wall art, 0.08 mm, 100 % infill, lowest speed, ~80 % of the bed, ~6 h; multicolour Benchy ~2 h; "Little Buddha" ~35 min; Elegoo name tag (2 colours) ~1 h. Notebookcheck: average colour change ≈ 1 min 10 s (74 and 78 colour changes on two test models)
2 sources
First-layer consistency reportsLeaning positive with one specific measured defect. Notebookcheck: "The default z offset is a bit too low resulting in wavy first layers, but lowering the bed by 0.025 mm fixes that." Against that: CreativeBloq "Bed adhesion was excellent, using the textured side of the plate, but not so strong that removal posed any issues"; Tom's reports no first-layer problems across six materials but lists "a cold plate that could use more grip" as a quirk; Notebookcheck also needed a 10-minute 110 °C bed preheat for ASA adhesion
1 source
Review scores from major outlets (list outlet + score + date)TechRadar Pro 4.25/5 (Alastair Jennings, 2026-01-28) — sub-scores Value 5, Design 4, Features 4.5, Performance 4.5 (these four average 4.5, not the 4.25 TechRadar publishes as its headline — a second arithmetic inconsistency in the same review, alongside the "29 out of 30" problem in E5). Tom's Hardware 4/5 (Denise Bertacchi, 2026-02-18). CreativeBloq 9/10 (Rob Redman). 3DwithUS 4.4/5 (Phil Macey, 2026-02-09). GamesRadar+ 5/5 (Benjamin Abbott). Notebookcheck — reviewed 2026-03-02, no numeric score extracted; verdict "Solid entryway into the world of multi-color printing … the hardware is on point but the software needs more tuning". Hackster.io (2026-01-26) and VoxelMatters — positive, unscored
1 source
Read with context

Sources & notes

How to read this guide

This guide presents the catalog’s sourced record for this exact model. Source dates belong to each reference; a page update does not mean every specification has been rechecked.

Prices are dated listing snapshots, with configuration, regional and sale conditions. They are not current offers. Unknown values remain unfilled.

Firmware, variants and available accessories can change. Read the original reference before making a purchase or planning a process.

Catalog references126 sources
Ownership research & coverage

Firmware, modifications and troubleshooting links are checked against the model and stated hardware or firmware revision. A creator’s compatibility claim is not a fit test by this site. Official troubleshooting describes a diagnostic route; individual reports do not establish a failure rate.

This review did not verify firmware customization or update guidance, printable modifications for this exact model. That is a research gap, not a claim that none exist.

Compare Elegoo Centauri Carbon 2 with another printer