Creality K1 Stringing Fix: Retraction, Heat and Drying

The Creality K1 stringing fix that actually sticks is not “turn retraction up”. The K1 is a direct-drive CoreXY rated for 600 mm/s, 20000 mm/s² and a 32 mm³/s hotend, per the Creality K1 product page, and on a machine this fast the usual causes of hair-fine strands are wet filament, a nozzle 10°C too hot for the spool, and a retraction profile tuned for a different filament. Retraction length is the last knob, not the first. Here are the checks in the order that costs the least time, with exact values for PLA and PETG on the stock 0.4 mm nozzle.

What K1 stringing is, and what it is not

Stringing is molten plastic that keeps flowing out of the nozzle during a travel move, leaving a strand between two islands of the part. The Prusa Knowledge Base puts the causes in the right order: too much heat, retraction not tuned, and filament that has absorbed moisture.

Two things look like stringing and are not:

• Fuzz and popping. Frizzy rather than clean strands, a rough matte surface on otherwise glossy PETG, crackling at the nozzle. That is steam. No retraction setting fixes it.
• Blobs and zits at the start of each perimeter. That is melt-zone pressure, which on the K1 belongs to pressure advance, not retraction length.

The K1 ships with a 0.4 mm nozzle, a direct drive extruder and a hotend rated to 300°C, per the Tom’s Hardware review, which found the pre-sliced Hyper PLA Benchy “smooth and string free”. A stock K1 on the stock profile with dry filament should not string. If yours does, something in that sentence changed.

Step 1: Rule out wet filament before touching a setting

The Creality filament drying guide says to dry PLA and PETG before use “when the ambient humidity exceeds 50%”, and shows PETG extruding with bubbles as the tell for absorbed moisture. Its drying table for a Creality filament dryer:

Filament	Dry temp	Dry time
PLA, PLA-CF	55°C	8 to 12 h
PETG	60°C	8 to 12 h
ABS, ASA	65°C	8 to 12 h
TPU	70°C	8 to 12 h

TPU and PA are listed as “must” dry rather than “suggest”, and a spool that has sat open through a humid month is a wet spool whatever the label says.

Without a dryer, use a food dehydrator: CNC Kitchen’s comparison of drying methods found it removed roughly twice as much moisture from PLA and PETG as a vacuum chamber, and that heat drying “removes more moisture in total”. Silica gel keeps a dry spool dry; it does not dry a wet one. Method details are in 3D Filament Guide’s drying walkthrough.

Quick check: heat the nozzle to print temperature and extrude 50 mm from the screen. If the strand bubbles, hisses, or comes out rough, dry the spool and reprint before you change anything in the slicer.

Step 2: Retraction settings for the K1

Start from what the machine profile actually says. The OrcaSlicer Creality K1 (0.4 nozzle) profile ships with:

• Retraction length: 0.8 mm
• Retraction speed: 40 mm/s, deretraction (prime) speed: 40 mm/s
• Z hop: 0.4 mm, type Auto Lift
• Wipe: on, retract before wipe: 70%
• Retraction minimum travel: 1 mm

Those are direct-drive numbers and the correct neighbourhood. The OrcaSlicer retraction test docs say that for PLA and ABS “setting the retraction length to 0.2mm - 0.4mm should suffice” once everything else is right, and that if stringing persists at the top of a tower, the fix is to “dry your filament and ensure that your nozzle is properly installed without any leaks”, not to keep adding length. Prusa caps retraction on its own direct-drive machines at 2 mm; treat that as the ceiling on the K1 too. Past it you pull molten plastic into the heatbreak and trade strings for clogs.

Concrete starting points for the K1, stock 0.4 mm nozzle, 0.2 mm layers:

Filament	Nozzle	Retraction	Speed	Z hop	Wipe
PLA	210 to 220°C	0.8 mm	40 mm/s	0.4 mm	on
PETG	235 to 245°C	1.0 to 1.2 mm	40 mm/s	0 to 0.2 mm	on

The PETG row is the one people get wrong. PETG oozes more than PLA and, like the flexible filaments Prusa says “usually need longer retractions”, wants the long end of the tower. It also tends to lift with the nozzle on a Z hop and leave a strand at the top, so print the tower once at 0.4 mm and once at 0 and compare.

Creality Print exposes the same controls under its extruder tab; the Creality Print parameter reference describes each, and notes that travel through infill or from support to support “will generally not retract”, so strings inside infill are normal and strings across open air between walls are not.

Step 3: Nozzle temperature

If the tower is clean at 1.0 mm but the part still strings, the plastic is too fluid. Prusa’s advice is to “decrease the nozzle temperature by 5 to 10°C and check whether there’s less stringing”. The K1 hotend heats fast (the product page claims 200°C in 40 s), so temperature towers are cheap. Run the OrcaSlicer temperature tower for PLA from 230°C down to 190°C in 5°C steps and pick the coolest block with clean bridges and no layer separation; the docs advise the middle of the acceptable range, or the high end if you are pushing speed and flow. On the K1’s fast profiles that usually lands PLA at 210 to 220°C, because the extra heat is what keeps 32 mm³/s flowing.

Step 4: Travel, minimum travel, and pressure advance

Three settings interact with retraction and are worth checking once.

Minimum travel. The K1 profile retracts on any travel over 1 mm. Prusa notes that raising this “will get lower printing times, but increase oozing and stringing”. Leave it at 1 mm for small detailed parts.

Wipe. Retract-before-wipe at 70% means most of the retraction happens while the nozzle is still dragging along the perimeter it just finished, which removes the bead that would become a string. Turning wipe off to save time is a common cause of new stringing after a profile tweak.

Pressure advance. The K1 runs Klipper pressure advance, which per the Klipper docs “reduces ooze during non-extrude moves” by dropping melt-zone pressure before the head decelerates. The same page still recommends “a small retract value in the slicer (eg, 0.75mm)” alongside it, which is where the K1’s 0.8 mm default sits. If you have edited printer.cfg or replaced the hotend, retune PA with that page’s tuning tower (TUNING_TOWER COMMAND=SET_PRESSURE_ADVANCE PARAMETER=ADVANCE START=0 FACTOR=.005 for direct drive) before blaming retraction; PA too low reads as corner blobs plus strings on exits.

A nozzle swapped and not torqued hot leaks around the heatbreak thread, and that leak shows up as stringing no slicer setting touches. If the strings started right after a nozzle change, start with the nozzle and heat creep guide.

How to verify

Use the built-in test rather than a random model, so the variable is isolated.

1. In OrcaSlicer, Calibration, Retraction test. Use the direct-drive defaults: start 0 mm, end 2 mm, step 0.1 mm. Filament dried, temperature tower already done.
2. Print it. Find the lowest section with no strands, measure its height with calipers, and read the matching length from the Calib_Retraction_tower comments in the G-code preview. The docs are explicit: the answer is “the shortest one that produces the cleanest tower”.
3. Enter that length in the filament profile, not the machine profile, so PETG and PLA keep their own values.
4. Confirm on a two-pillar stringing test at real print speed.

Success criteria: no strand between pillars that survives a fingernail flick, no blob where the nozzle re-enters the wall, and no gaps in the first two perimeters after a travel (retracted too far or primed too little). If the tower is clean at every length but the real part strings, the difference is speed and heat, and Step 3 holds the answer. The K1 Max and K1C use the same procedure and starting values; hardware differences are in the K1 Max vs K1C comparison.

Related across the network

• How to Fix Stringing in Cura: Retraction, Temperature, Travel — slicerguide.com
• Stringing and Oozing: Causes and How to Tune Retraction — fdmdesk.com
• How to Fix Stringing on 3D Prints: Retraction Settings — printlabguide.com
• How to Dry Filament: Temperatures by Material — 3dfilamentguide.com
• Best Cura Settings for PETG: Exact Values and Why They Matter — slicerguide.com