Pressure Advance / Linear Advance

TL;DR

Use Pressure Advance (Klipper) / Linear Advance (Marlin) to clean up corner bulges and seam blobs at higher acceleration. Print a single-wall sharp-corner test at fixed speed and acceleration. Increase the value until bulges disappear, then stop before corners start to look thin or gappy.

What Pressure Advance is fixing (the physical picture)

Molten plastic in your hotend acts like a pressurized, springy column. When the toolhead accelerates, pressure takes a moment to build, so the nozzle under-extrudes briefly. When the toolhead decelerates, stored pressure keeps pushing, so the nozzle over-extrudes briefly. Pressure Advance / Linear Advance adds extrusion early during acceleration and pulls it back early during deceleration so line width stays even through speed changes, corners, and the seam.

Where it helps most (and where it can’t)

• Sharp corners and small features with constant accel/decel
• Visible seam start/stop where blobs usually form
• Higher speed, higher acceleration, and higher jerk (where pressure lag is bigger)
• Blobs that persist even after reasonable retraction tuning (it complements retraction; it doesn’t replace it)
• Not a fix for wrong e-steps/rotation distance, wrong flow/extrusion multiplier, partial clogs, extruder slip, wet filament, or unstable nozzle temperature; those must be solved first

Quick setup checklist (so your test means something)

• Verify what your firmware calls it (Marlin: Linear Advance K; Klipper: pressure_advance) and that it’s enabled
• Use one filament and keep it dry; note filament, nozzle size, and hotend/extruder setup
• Hold speed, acceleration, and cooling constant for the whole test
• Keep line width, layer height, temperature, and flow constant across all sections
• Disable confounding tweaks (no per-section flow overrides; keep slicer settings consistent)

Run a practical calibration (simple and repeatable)

Print a test that forces repeated acceleration/deceleration through sharp corners, ideally a single-wall pattern so width changes are obvious. Either print multiple small parts at different values, or use one test that changes the value by section. Change only the advance value between sections. If you also change temperature, speed, acceleration, or flow, you won’t know what actually improved or broke the result.

Pick the value by what you see on the wall

• Too low: corners bulge, outside corners look overfilled, seam start/stop blobs, line gets wider at slowdowns
• Too high: corners look pinched, underfill at/after corners, tiny gaps right after deceleration, seam start can look starved
• Just right: straight walls stay the same width, corners are crisp without swelling, seam is least noticeable without new gaps

Validate on a real part (and when to retune)

After you pick a value, print a small real part with a visible seam and sharp corners at your normal speed and acceleration. Recheck (and often retune) when you change filament type, nozzle size/material, hotend/extruder hardware, or you significantly change acceleration or print speed—each change alters how pressure builds and releases in the melt path.