Flow and Temperature Calibration, Done Properly

Most “bad slicer settings” problems are actually uncalibrated-printer problems. If flow rate and temperature are wrong, no quality setting can compensate — you’re tuning seam placement on a printer that’s over-extruding by 6%, which is like adjusting the mirrors on a car with a bent axle.

Two calibrations underpin everything else: flow rate (how much plastic comes out) and temperature (how well it bonds). Get these right, in this order, before trusting any other setting.

Why order matters

Run flow calibration first, temperature second. Here’s why: a temperature tower judges layer adhesion and surface quality, and both are skewed if flow is wrong. Over-extrusion makes every temperature on the tower look rough; under-extrusion makes every band look weak. Calibrate flow first so the temperature tower is judging temperature, not flow error.

Then re-check flow once at your final temperature, because flow ratio shifts slightly with hotend temperature. One pass each, in order, is enough for most filaments.

Step 1: Flow ratio (extrusion multiplier)

Flow ratio scales how much filament the extruder pushes. The goal is a top surface that’s neither gappy (under-extruded) nor bulging and rough (over-extruded).

The modern method (OrcaSlicer)

OrcaSlicer’s calibration menu generates a flow-ratio test as two passes:

1. Pass 1 (coarse): prints several thin-walled patches at flow modifiers from roughly -10% to +10%. Pick the patch with the smoothest, gap-free top.
2. Pass 2 (fine): narrows the range around your pass-1 winner in ~1% steps. Pick again.

Enter the resulting flow ratio into the filament profile. This two-pass approach converges faster and more reliably than eyeballing a single calibration cube.

The classic method (PrusaSlicer / Bambu Studio)

Neither ships an identical built-in two-pass test, so the durable approach is the thin-wall measurement:

1. Print a single-wall open box (a 0.4 mm-wall cube with no top/bottom) at 100% flow.
2. Measure the wall thickness with calipers at several points and average it.
3. New flow = current flow × (intended line width ÷ measured wall thickness).

Measure-driven beats eyeball every time. Calibration is a measurement discipline, not a vibe.

Reading the result

• Gaps between top-surface lines, visible pinholing: under-extruding — raise flow.
• Raised ridges, rough/bumpy top, dimensions oversized: over-extruding — lower flow.
• Smooth, fused, flat top with correct dimensions: correct.

Flow ratio is per filament type and brand, sometimes per spool. A new brand of PLA is a new calibration.

Step 2: Temperature tower

A temperature tower prints the same geometry in stacked segments, each at a different hotend temperature, typically stepping in 5 °C increments across the material’s range.

Generating it

• OrcaSlicer: Calibration → Temperature. It generates the tower model and inserts the per-segment temperature G-code automatically. Set the start/end temperatures to bracket the filament’s printed range.
• PrusaSlicer / Bambu Studio: import a temperature-tower STL and use the per-height temperature change (a “Custom G-code at layer/height” entry, or the height-based color/temperature change). The slicer inserts an M104/M109 at each band.

Always start from the temperature range printed on the spool. A “PLA tower” from 190 °C to 230 °C is pointless if that spool’s range is 200–220 °C — test 195–225 to bracket it slightly, not arbitrary numbers.

Judging the tower

Inspect each band for four things:

1. Layer adhesion — try to snap or peel a band. Higher temperatures usually bond stronger. This is the most important criterion for functional parts.
2. Stringing / oozing — wisps between the tower’s overhang fingers or pillars. Lower temperatures string less.
3. Bridging and overhang quality — sag on the test’s overhang features. Lower temperatures hold overhangs better; too low and adhesion suffers.
4. Surface finish — blobbing, scarring, or zits worsen at temperature extremes.

The right temperature is the lowest one that still gives strong layer adhesion with acceptable stringing and overhangs — usually a single value that balances all four, not the absolute hottest or coolest band.

Step 3: Re-verify flow at the chosen temperature

Flow ratio drifts a little with temperature. After picking a temperature from the tower, run one more fine flow pass at that temperature and lock both values into the filament profile together. Treat flow + temperature as a paired calibration stored per filament.

Common mistakes

• Calibrating on the wrong surface. Judge flow on a top surface or a single-wall thickness, not on infill or side walls — those hide the error.
• Towering outside the spool’s range. You’ll pick a temperature that looks fine on the tower but degrades the polymer or under-bonds in real prints.
• One global flow number for all filaments. Flow is per material and brand. A profile that’s perfect for one PLA can over-extrude another by several percent.
• Skipping straight to speed/quality tuning. Pressure advance, max volumetric speed, and retraction tests assume flow and temperature are already correct. Calibrate in order.

What to calibrate after this

With flow and temperature locked, the next tests — pressure advance (crisp corners and seams), retraction (stringing), and max volumetric speed (the real speed ceiling) — are meaningful. Run those before chasing high print speeds; the core settings guide explains why volumetric speed, not the speed number, is the true cap.

Where to go next

For applying calibrated profiles per material, see the material profiles guide. For diagnosing defects that survive correct calibration, see troubleshooting print defects in the slicer.

For the measurement and materials-science background behind these tests, PrintLabGuide ↗ goes deeper, and FDM Desk ↗ covers printer-specific calibration profiles.