Reduce Stringing Without Hiding the Real Cause

The mistake I made in my first year of running printers was changing three or four things at once the moment stringing showed up, a bit of retraction, a bit of temperature, sometimes the whole travel speed, all in the same slicer session. Sometimes the strings went away. I never knew why, and a few spools later they’d be back, because I hadn’t actually fixed anything, I’d just landed on a combination that happened to cancel out one particular batch of damp PETG. What works is slower: change one variable, print the same small test piece, look closely at what came out, then decide whether to keep going. It takes longer than cranking retraction up by 2 mm and hoping, but it tells you which variable was responsible, and that’s what lets you trust the profile you save.

1. Check the filament first

Before you open the slicer, listen to the printer. Wet filament tends to announce itself right at the nozzle: small pops or cracks as trapped moisture flashes to steam, tiny wisps that look like steam rising off the melt zone, and a strand that comes out rough instead of glassy smooth. Pull one of the strings off a recent print and roll it between your fingers. Dry filament strings feel smooth and snap apart cleanly. Wet filament strings often feel slightly rough or bumpy, sometimes with tiny bubbles you can spot if you hold the strand up to a light.

PETG, nylon, and TPU pull moisture out of the air fast enough that a spool left open on a shelf in a humid room for a few weeks can pick up enough water to cause this. PETG is the one that keeps stringing even after you have done everything on this page, for reasons that come down to the material rather than your settings; Fix PETG Stringing covers what changes when that is the filament you are running. PLA is more forgiving but not immune, especially once a spool has sat open for months. If you’ve got a second spool of the same material that’s stayed sealed with desiccant, print the same test piece on both and compare. If not, dry the suspect spool instead. A dedicated filament dryer, the kind of small heated box brands like SUNLU, Creality, or eibos sell, holds temperature far more evenly than a kitchen oven, and follows the time and temperature the filament maker lists for that specific material, since nylon and PETG dry at different settings entirely. Skip the oven unless its thermostat actually holds steady: home ovens can swing 10 to 20 degrees around their setpoint, and some spools are made of a plastic that warps well below the temperature the filament itself needs to dry.

2. Test a lower nozzle temperature

Once the filament itself checks out dry, print the same small test piece again, but drop the nozzle temperature by 5 °C from where you’d normally run that material. A simple two-post or two-wall tower with a gap between the features shows this better than a full model, because it makes it obvious exactly how much material bridges the gap. Hold the result up to a lamp or a window afterward. Backlighting turns even faint strings into bright thin lines you’d otherwise miss under ordinary room light.

Lower temperature helps because there’s simply less molten plastic sitting at the nozzle tip when it crosses a gap, so less has the chance to ooze out and stretch into a string. But every filament has a working range for a reason. PLA typically runs somewhere in the 190 to 220 °C band and PETG higher, often 230 to 250 °C, and the low end exists because layers need enough heat to actually fuse together. Check bonding on your test piece, not just the strings. If you can peel or snap layers apart with your fingers, or the walls feel weaker than they should, you’ve dropped below the point where this filament bonds properly. Treat this test as a ceiling you’re feeling for, not a race to the lowest number the printer will accept.

3. Adjust retraction in small steps

Retraction is the setting most people reach for first, and it’s often the right one, just not in big jumps. How far you should retract depends heavily on what kind of extruder is pulling the filament. Direct-drive setups have the gear right at the hot end with almost no slack in the path, so they typically only need something like 0.5 to 2 mm. Bowden setups have several centimeters of tubing between the gear and the nozzle that can compress slightly under pressure, so they commonly need more, often 3 to 7 mm. Treat your printer’s stock profile as the starting point rather than a guess, and move away from it in small steps, commonly around 0.2 mm at a time for direct drive and a bit more for Bowden.

You’ll usually know retraction has gone too far before you even look at the print. Grinding filament makes a distinct sound, a rhythmic clicking or chattering as the extruder gear skips against the filament instead of biting into it cleanly, and afterward you’ll often find a flat worn patch or a bit of fine plastic dust near the gear. Push retraction distance or speed too far and you can also run into heat creep, where filament just above the melt zone gets pulled back and forth so often that it softens and swells, then jams the next time it tries to feed forward. That shows up as a stutter or a skipped click from the extruder motor a few layers later, followed by thin, starved extrusion. None of that means retraction is the wrong tool. It means you’ve found the edge of what your hardware can do, and the setting that works is one step back from there.

4. Review travel behavior

Travel moves are the ones where the nozzle isn’t printing, just repositioning between one feature and the next, and they’re where most visible strings actually form. Most slicers have a setting for this: Cura calls it Combing Mode, PrusaSlicer and OrcaSlicer call it Avoid Crossing Perimeters. It routes travel moves through areas about to be covered by more plastic anyway, so any ooze along the way gets buried instead of left hanging in open air. It works well on solid parts, but it can add real time to a print, sometimes a noticeable amount, because the nozzle is no longer taking the shortest path between two points.

Raising travel speed helps in a different way: less time crossing a gap means less time for material to ooze, even without changing anything else. Older, lighter-framed printers often top out somewhere around 120 to 150 mm/s before things get rough, while newer machines built around stiffer frames and lighter toolheads can push well past 300 mm/s. Go past what your specific frame and motion system can handle smoothly, though, and you’ll start seeing ringing, faint wavy echoes in the plastic near corners where the frame flexed like a tuning fork, or you’ll hear it rattle on fast direction changes. Some slicers also offer a small Z hop on travel moves, lifting the nozzle clear of the last layer before it moves, which helps on tall, thin features but adds wear to the Z axis and doesn’t touch why the ooze is there in the first place. Change routing or speed only after moisture and temperature are already sorted, otherwise you won’t know which change actually helped.

Before saving the profile

Before any of this becomes a saved profile, run back through what actually changed and why it worked, not just whether the strings are gone.

• Filament condition is known, not assumed.
• The same test piece was used for every step.
• Layer bonding is still strong, not just the strings gone.
• Extrusion resumes immediately after travel, with no gaps or starved lines.
• Every changed value is still within the printer profile’s safe range.