Start chasing PETG strings with a small test you can repeat. It gives you somewhere to compare material condition, temperature and retraction without committing a whole production print to each adjustment. Keep the part's job in mind as you work: the hairs alone tell you very little about whether its layer adhesion is good enough.
Establish the starting conditions
Record filament identity, nozzle, temperature, retraction and travel settings. Inspect the filament path and nozzle using the printer's maintenance instructions.
Prusa's stringing guidance[1] works through temperature, retraction, travel and material condition. Use it to organize the checks, while keeping its machine-specific examples tied to the hardware they describe.
Address moisture with approved equipment
If the filament maker recommends drying, use its time and temperature for the exact product and a dryer suitable for both filament and spool. Keep the dried material in the recommended storage conditions.
Use drying equipment whose specifications cover the material’s required temperature and the spool it sits on, following the equipment’s instructions. Prusa’s drying guide[2] explains why filament and spool limits both matter. I would choose that controlled setup over an improvised cardboard enclosure on a heated bed.
Set temperature before fine retraction changes
Test within the filament and hotend limits, comparing each sample for extrusion consistency and surface quality as well as stringing. The temperature you keep also needs to suit the strength required by the part. Fewer hairs would be a disappointing trade if the rest of the print no longer met the job.
Once you've accepted a temperature, hold it steady while testing retraction. Start from the printer's supported profile and change distance or speed separately so the comparison remains useful. Bowden and direct-drive examples need to be read in the context of their hotend design before you borrow a setting.
Review travel and wipe settings
Follow the moves between islands in the slicer's preview, then test the supported travel-avoidance or wipe controls within the machine's motion limits. Leave attempts to snap strings by raising cornering settings out of this comparison: Marlin jerk and Klipper square-corner velocity are different controls, and their numbers cannot be swapped between profiles.
Avoid stacking pressure-control changes blindly. If you change a wipe, coasting or pressure-advance setting, record it and recheck seams for missing material.
Choose a repeatable result
Put the samples under the same lighting and look closely at the restart points as well as the hairs. Judge extrusion and strength on their own merits alongside that surface comparison; neither a string-free finish nor a web of strings settles those questions.
Save the profile that meets the part's requirements and try it on representative geometry before committing the production print. If only minor strings remain, let the part cool and remove them with a suitable mechanical method. A heat gun brings a risk of distorting the print, so it deserves more thought than a quick finishing blast.
