Tune a new filament spool: temp tower, flow, retraction, volumetric speed

The new spool arrives. The label says “PETG 230–250 °C” and that is all. The slicer profile is a safe middle value the printer maker chose so that anything prints, not the best value for this spool on your machine. This lesson is a four-test routine that takes about an evening and gives you a profile for the whole spool.

Numbers assume an FDM printer with a 0.4 mm nozzle and 0.2 mm layers. Setting names follow OrcaSlicer and PrusaSlicer; Bambu Studio uses almost the same names as Orca.

Before you start: three things to have ready

• Dry filament. A damp spool skews every test below. If the spool has been open for a while, dry it first (see the next lesson).
• A clean, tight nozzle. Residue around the nozzle causes strings no setting will fix.
• A calibrated extruder (E-steps / rotation distance). Modern printers ship with this correct; leave it alone.

Why the order matters

The OrcaSlicer calibration guide recommends: temperature → max volumetric speed → pressure advance → flow ratio → retraction. Temperature sets how runny the plastic is; every other value is measured on top of that, so changing it later means measuring everything again. This lesson skips pressure advance: many newer printers calibrate it in firmware, and the speed lesson covers it.

Change one thing at a time. If you change two and the print improves, you do not know which one helped.

Step 1: temperature tower

A tower where each block prints at a different temperature, dropping 5 °C per block from top to bottom. Common starting ranges:

Material	Range to test
PLA	190–230 °C
PETG	220–260 °C

For other materials, take the spool label’s range and test 5–10 °C wider on each end.

In OrcaSlicer: Calibration → Temperature, enter start/end/step, and the tower comes with the temperature-change G-code built in. PrusaSlicer cannot set temperature per height range, so you add M104/M109 lines in “Before layer change G-code” yourself, or load a tower from Printables that includes them.

Judge each block on stringing, layer adhesion, overhangs, bridges and surface. Rough rule: cooler reduces strings but weakens bonding; hotter bonds better but overhangs sag. If several blocks look equally good, take the middle, or lean high if you plan to print fast.

Step 2: max volumetric speed

The most plastic the hotend can melt per second, in mm³/s. The slicer uses it as a ceiling and slows down wherever a line is thick or wide enough to exceed it. Reference values from the Prusa Knowledge Base:

Material	Typical range (mm³/s)
PLA	12–20
PETG	8–15
ABS / ASA	10–16
Nylon (PA)	8–12
PC	6–10
TPU	2–5

The hotend sets a second limit: a standard hotend such as an E3D V6 manages about 8–12 mm³/s, high-flow nozzles (CHT) 20–35 mm³/s. The lower of material and hotend applies.

The test in OrcaSlicer: Calibration → Max flowrate, defaults start 5, end 20, step 0.5. It prints a part whose flow rises with height. Find the height where the surface starts to fail (broken lines, a change in sheen) and measure it with calipers.

Example: the surface fails at 19 mm.

1. Raw value = 5 + (19 × 0.5) = 14.5 mm³/s
2. The guide recommends taking 10–20% off, because the test has no retractions and no tight corners: 14.5 × 0.85 ≈ 12.3 → set 12 mm³/s
3. The real top speed for a 0.45 mm line at 0.2 mm layers = 12 ÷ (0.45 × 0.2) ≈ 133 mm/s. Set anything faster and the slicer trims it for you.

Enter it under Filament settings → Max volumetric speed. PrusaSlicer has it twice (Print Settings → Speed for the hotend, Filament Settings → Advanced for the material); the lower one wins.

Step 3: flow ratio (extrusion multiplier)

A multiplier on how much plastic is pushed out: Extrusion multiplier in PrusaSlicer, Flow ratio in Orca and Bambu. Default 1.0; tuned values usually sit between 0.9 and 1.1. Outside that, suspect a partial clog or wrong E-steps before touching the number.

The measured method (Prusa): print a cube in vase mode (single wall), measure the wall thickness mid-face at three or more points and average them.

Example: the profile’s extrusion width is 0.45 mm and the wall measures 0.46, 0.48 and 0.47 → average 0.47 mm.

• New extrusion multiplier = 0.45 ÷ 0.47 ≈ 0.957
• Reprint and measure again; the wall should now be close to 0.45.

Cheap digital calipers are not precise enough for a single wall; if in doubt, use the visual method.

The visual method (OrcaSlicer YOLO): Calibration → Flow rate → YOLO prints 11 blocks, each ±0.01 apart. Pick the one with the smoothest top, no gaps between lines and no raised ridges, then add that block’s value to your current flow, for example 0.98 + 0.01 = 0.99. The older two-pass method uses old × (100 + modifier) ÷ 100.

How wrong looks: too much flow → rough top, plastic piling up at the edges, parts that no longer fit. Too little → visible gaps between lines or walls.

Step 4: retraction

Pulling the filament back while the nozzle crosses a gap so plastic does not ooze into a string. Length and speed are the values that matter.

OrcaSlicer: Calibration → Retraction test prints a tower where each section uses a different retraction length.

Extruder type	start	end	step
Direct drive	0 mm	2 mm	0.1 mm
Bowden	1 mm	6 mm	0.2 mm

Find the cleanest section, measure its height, and search the G-code preview for Calib_Retraction_tower to read the matching length. Choose the shortest clean value, not the largest: pulling too far drags molten plastic into the cold zone and clogs. PLA and ABS sometimes need only 0.2–0.4 mm; flexible filament (TPU) usually needs more.

If the whole tower still strings heavily: dry the filament, check the nozzle for leaks, lower the temperature in 5–10 °C steps and test again.

Last step: save it as a profile

Save the filament profile under a name that identifies the spool, such as PETG <brand> black 0.4, and put the summary in its notes field. From this lesson’s tests:

Setting	Before	After
Nozzle temperature	245 °C (middle of label)	240 °C
Max volumetric speed	15 mm³/s (default)	12 mm³/s
Flow ratio	1.00	0.957
Retraction length	0.8 mm	0.6 mm

These are a worked example, not values for every printer. A different colour or batch can shift flow and temperature a little; check with a small test part before a long print.

A good value is one that prints the same every time, not the prettiest result once. If you have to re-tune every spool, another variable (moisture, a worn nozzle) is not under control yet.

Common mistakes

• Tuning retraction before temperature. Change the temperature later and the strings come back.
• Using the raw max volumetric speed from the tower without taking 10–20% off. Real parts have retractions and tight corners, and the extruder falls behind in patches.
• Fixing flow to hide another problem. A flow of 0.85 means something is wrong, usually a half-clogged nozzle or damp filament.
• Forgetting to leave calibration mode. In OrcaSlicer, start a new project afterwards or the next model inherits the test G-code.
• Testing with a damp spool. You get values that work that day only; after drying, everything shifts.

Safety

The towers print quickly, but the nozzle is at 190–260 °C; let it cool before pulling parts off. Test ABS/ASA in a ventilated space, and stay near the printer during the upper end of the volumetric speed test, where a clog is most likely.

Next: Damp filament: drying and storage

Sources

• OrcaSlicer Wiki: Calibration Guide (test order)
• OrcaSlicer Wiki: Max Volumetric Speed calibration
• OrcaSlicer Wiki: Retraction calibration
• Prusa Knowledge Base: Extrusion multiplier calibration
• Prusa Knowledge Base: Max volumetric speed