Direct answer: Fix a rough PETG top surface by reading the cooled texture before changing settings. Broad sagging or pillowing points first to weak support below the roof. Lines that remain open through the final top layer point to insufficient total top thickness or unstable extrusion. Random pits, bubbles, or a sudden decline after storage support a spool-condition check. Raised ridges, glossy smears, scratches, or dragged deposits move heat, excess material, and nozzle contact higher on the list.
This guide covers one exact failure: the final upward-facing PETG skin is rough, ridged, pitted, smeared, or incompletely closed. It does not cover a rough support-facing underside, open-air stringing, a first layer that will not stick, or an entire part with weak layers.
Editorial scope: This workflow uses current manufacturer guidance and does not claim hands-on testing. PETG formulations, printers, nozzles, profiles, and cooling systems vary. Keep temperatures, fan changes, drying, and machine service inside the exact printer and filament makers' supported limits.
Disclosure: GoodPrints may earn from qualifying purchases through the clearly labeled affiliate link on this page, at no added cost to you.
Confirm the defect is a PETG top-skin failure
Let the part cool and inspect the final horizontal face under side light. Mark whether the defect is a low area, open line, pit, raised ridge, glossy smear, scratch, or deposited blob. Then open the slicer's layer preview and compare that location with the infill below it, every top-solid pass, travel moves, and any direction change.
Separate four lookalike failures
If every extrusion line is thin or intermittent, use the under-extrusion workflow. If hairs span open travel, use the PETG stringing guide. If the nozzle is scraping walls or infill as well as the roof, use the nozzle-scraping diagnostic. If only the ironed pass is rough, disable ironing and prove the un-ironed roof first.
Match the cooled PETG pattern to the first useful check
| What you observe | Leading hypothesis | First clean proof |
|---|---|---|
| Broad low waves or pillowed patches above infill | Weak support below the roof | Inspect the layer below the first top-solid pass and shorten one unsupported span |
| First roof pass is rough and later passes improve, but the final face stays open | Insufficient total top thickness | Add one controlled thickness step without changing infill, flow, heat, or speed |
| Random pits, bubbles, popping, fuzzy lines, or a spool that worsened after exposure | Spool-condition drift | Repeat the same roof before and after the maker-approved drying process |
| Glossy smears or soft lines pushed sideways mainly on broad top areas | Heat, cooling, or top-speed mismatch | Test one supported temperature, fan, or top-speed step while holding the others fixed |
| Raised crowded ridges plus heavy dimensions or overfilled solid regions | Excess extrusion or pressure | Measure a known feature and compare walls before changing flow |
| Scratches, dark streaks, or sticky PETG dragged over otherwise sound lines | Nozzle contact or exterior residue | Inspect the toolpath before the mark and the cooled nozzle exterior |
Find the first bad PETG roof layer before changing settings
The cooled final surface can hide where the failure started. Open the exact saved project in layer preview and inspect four points in order: the infill or support directly below the roof, the first solid PETG pass over it, the intermediate solid passes, and the final visible top pass. Then compare that sequence with the failed part and record the first layer and location that becomes abnormal. This keeps a support problem, a short roof stack, unstable extrusion, a bad final pass, and nozzle contact from being treated as the same defect.
| First abnormal PETG evidence | Leading cause branch | One controlled next check |
|---|---|---|
| The first roof pass sags over the same wide infill cells, and every later pass inherits those low spots | Support below the roof | Change only the supported span, infill pattern or density, or local geometry and repeat the same roof |
| Each solid pass closes more of the same openings, but the production roof ends before it becomes continuous | Total top thickness | Add one controlled thickness step without changing infill, PETG condition, heat, fan, flow, or speed |
| Random skips, pits, bubbles, or thin lines move between passes, and walls, infill, or a steady purge also vary | Spool condition, feed, or extrusion stability | Prove dry material, a steady spool path, a clean purge, and a clear nozzle before editing roof-only controls |
| The lower solid passes are closed, but the final top pass piles up, smears, stays glossy-soft, or leaves parallel ridges | Final-pass heat, speed, cooling, or proven flow | Test one modest maker-supported change while support and shell thickness stay fixed |
| The roof is sound until one raised PETG line curls or is struck, after which scratches and carried residue appear | Nozzle contact with a local trigger | Correct the curl, excess material, residue, warping, or mechanical contact before using a cosmetic workaround |
| The normal top pass is continuous and only the ironing pass becomes rough | Ironing, not the roof itself | Disable ironing, confirm the unchanged roof, then tune ironing as a separate finishing operation |
If preview and the failed part still leave two plausible branches, print the same small roof coupon and change one variable at a time. Keep the nozzle, spool, layer height, orientation, infill, temperatures, cooling, and speeds fixed unless that item is the variable under test. Let PETG cool before comparing it under the same light, and repeat the winning setup once. A cosmetic coupon does not qualify a mating, sealing, loaded, pressure, electrical, heat-exposed, or safety-relevant part.
Check 1: restore the exact PETG baseline
Save the failing project and record the printer, nozzle material and diameter, layer height, PETG product, spool exposure history, material preset, total top thickness, infill pattern and density, top-surface speed, temperature, cooling, and flow-related overrides. Remove forgotten experimental overrides before comparing anything.
Use the maker's PETG starting point
Prusa's current PETG guide treats drying, cooling, temperature, adhesion, and stringing as related material controls. Use the correct manufacturer profile and guidance as the baseline. Do not import a popular PLA roof recipe or another machine's fan and temperature values.
Check 2: inspect the support immediately below the roof
The first top-solid lines bridge from the structure beneath them. If infill stops short of a wall, changes direction under the failure, or leaves a wide unsupported span, those first lines can sag. Every later top pass inherits that low spot, so the final PETG face can show waves or pillowing even when extrusion is otherwise healthy.
Prove support in layer preview
Stop on the layer below the first top-solid pass. Compare the failing area with a clean area. Prusa's current infill-pattern guide shows that patterns differ in direction, speed, strength, and support for upper layers. Test the smallest local change that gives the roof a shorter, more continuous span. Maximum infill is not the goal.
When rough PETG starts as bridge sag, use the bridge branch
Stop this roof-finish workflow when layer preview shows that the first solid pass crosses a true open gap rather than resting on infill or another supported surface. A drooping first bridge can leave every later top layer rough, but top-surface speed, ironing, and a blind global-flow change do not repair the original unsupported strand. Use the bridge-sag diagnostic for that first failure, then return here only if the bridge becomes stable and the later PETG top skin still remains rough.
| Preview and part evidence | Use this branch | Return condition |
|---|---|---|
| The first roof line hangs across open air and droops before later solid passes begin | Bridge-sag diagnosis | Return after the first bridge is attached and repeatable |
| The first roof pass rests on infill but sinks in the same wide cells | Roof-support and span check on this page | Continue here; it is not a free-air bridge |
| The bridge and lower roof are sound, but only the final visible pass is ridged, smeared, or scratched | PETG top-skin checks on this page | Start with the first matching heat, speed, flow, pattern, or contact evidence |
Change one bridge control without turning it into a global PETG recipe
Prusa's current poor-bridging guide says bridges work best over shorter distances and treats slicer settings, bridge speed, cooling, and bridge flow as separate controls. That is a bridge-diagnosis framework, not a universal PETG value. Keep the failing project, spool, nozzle, layer height, geometry, and ordinary top settings fixed. Inside the exact printer and filament makers' limits, test only one supported bridge-speed, bridge-cooling, or bridge-flow step at a time. Do not lower global flow or force maximum fan from one sagging span. Keep a change only when the first bridge attaches consistently, later roof passes close, and the cooled part does not gain thin walls, weak bonding, curl, or nozzle contact. A clean bridge coupon still does not qualify a loaded, sealing, pressure, heat-exposed, electrical, or safety-relevant part.
Check 3: verify total top thickness, not only layer count
A familiar layer count can become a physically thinner roof after a layer-height change. The first roof pass may be imperfect; the key evidence is whether later passes steadily close it. If the final pass remains open but every pass improves, one controlled top-thickness increase is a cleaner test than changing flow or heat.
Judge the whole roof stack
Use millimeters and layer count together. Prusa's current layers-and-perimeters guide documents top solid layers, top fill pattern, and the relationship between shell choices and the printed part. Keep the smallest top thickness that closes reliably over the actual support pattern. If added thickness simply repeats the same pits, weak lines, or ridges, return to baseline and continue.
Check 4: separate spool condition from feed restriction
Move moisture forward only when the evidence supports it: the same spool printed clean roofs before exposure, then developed popping, bubbles, random pits, fuzzy extrusion, or inconsistent gloss. Follow the exact filament maker's drying instructions and protect the spool during the repeat. A storage bag or dry box does not prove that a wet spool has been restored.
Compare the same slice before and after one proof
If maker-approved drying improves pits, popping, and roof consistency together, spool condition mattered. If walls, infill, or a steady purge remain thin or inconsistent, inspect spool drag, the feed path, extruder grip, and a partial restriction. Use the partial-clog guide when a purge is unstable after simple feed resistance is ruled out.
Check 5: test PETG heat, cooling, and top speed separately
PETG that stays soft can smear when the nozzle returns across a broad roof. A top pass that outruns clean deposition can leave uneven closure. Cooling changes how quickly lines firm up, but maximum fan is not a universal answer because geometry, bonding, enclosure state, formulation, and the printer's intended profile all matter.
Change only one process control
With support, thickness, and spool condition held stable, test one modest supported temperature step, one supported cooling step, or one top-surface speed reduction. Keep a change only if the cooled roof improves without weaker bonding, warping, dull starvation, or a new texture. If the part begins splitting on printed planes, move to the PETG layer-cracking diagnostic.
Check 6: distinguish excess material from nozzle drag
Raised lines can come from genuinely heavy extrusion, while scratches and sticky deposits can come from the nozzle touching material that rose for another reason. Compare dimensions, walls, infill, and a known calibration feature before lowering flow. Then inspect the cooled hotend exterior and the toolpath immediately before each mark using the printer maker's safe procedure.
Do not make residue the whole diagnosis
Cleaning the nozzle exterior can remove carried PETG, but it does not fix over-extrusion, a loose toolhead, a curled edge, a high crossing, or an internal restriction. Use the over-extrusion workflow only when measurements and other solid regions support excess material. Correct the contact mechanism separately.
Bambu Studio owners: calibrate flow only after the roof evidence points there
Bambu's current Flow Rate Calibration guide lists a rough or uneven surface as one possible sign of incorrect flow, but it also warns that not every defect comes from flow and places flow-rate work after other calibration checks. On this PETG symptom, enter that branch only when raised crowded lines, heavy dimensions, excess material in other solid regions, or repeatable gaps across more than the roof support it. Broad sagging above infill, exposure-linked pits, heat smears, and isolated nozzle scratches still belong to their earlier checks.
- Freeze the exact baseline. Save the printer, nozzle diameter, PETG product, filament preset, layer height, top thickness, temperature, cooling, speed, and the failing slice. Do not calibrate against a project that still contains unexplained overrides.
- Stabilize the spool first. Bambu's calibration guidance says the filament should be thoroughly dried and kept dry for reliable results. A changed moisture state can make the comparison ambiguous.
- Use the current supported calibration path. Follow the auto or manual workflow offered for the exact printer and Bambu Studio version. Select the sample with continuous, smooth lines rather than the one that merely hides gaps under crowded ridges, then save the result to the exact filament preset being tested.
- Reprint the unchanged roof. Keep support, shell thickness, heat, fan, speed, and geometry fixed. Keep the calibration only if two cooled repeats improve the roof and the other solid regions without making dimensions, walls, or layer bonding worse.
Bambu-specific limit: the current Bambu PETG usage guide publishes a 0.93-0.96 flow-ratio range for PETG-like filaments and warns that going below it can create under-extrusion and weaker layer adhesion while going above it can increase oozing or clogging. Treat that as a boundary for the guide's Bambu workflow, not a universal PETG setting and not permission to overwrite a third-party filament maker's supported profile. If the calibrated roof is smoother but the actual part loses fit, bonding, sealing, or repeatability, the change does not pass.
Check the top-fill pattern only after the PETG roof is closed
Use this branch only when layer preview and the cooled part show a continuous, supported roof with no random pits, open lines, broad sag, or nozzle strikes, but the remaining roughness repeats at the top-fill layout. A raised band where top lines meet the perimeter, directional sheen that changes with the line order, or buildup concentrated at starts and ends is pattern evidence. Roughness across every line, or a defect already visible on the first roof pass, belongs to an earlier check.
| What the closed PETG roof shows | Pattern test | Stop rule |
|---|---|---|
| The center is closed, but an excess-material ridge follows the perimeter meeting line | Compare the saved pattern with one maker-supported top pattern that changes perimeter overlap behavior | If the ridge stays in the same place, return to flow, temperature, contact, or geometry evidence |
| The surface is closed and sound, but line order creates inconsistent directional appearance | Run the same coupon with only the top-fill pattern changed | Do not accept extra seam marks or time cost merely to move the cosmetic defect |
| Pits, bubbles, gaps, sag, smears, or scratches continue through the roof | Do not change the pattern | Resume the first matching support, thickness, spool, process, or contact check |
Run one pattern A/B without hiding the original cause
Prusa's current infill-pattern guidance says Monotonic lines follows a consistent order while avoiding perimeter overlap, which can reduce excess material and produce a smoother-looking surface; it also warns that the pattern takes longer and can make seams visible. Treat that as a bounded slicer choice, not a universal PETG fix. Keep the model, orientation, nozzle, spool, layer height, roof thickness, infill, temperatures, cooling, speed, and flow unchanged. Compare cooled coupons under the same light and repeat the better result once. A pattern win can release a cosmetic face only after the earlier functional limits still pass; it does not qualify a leaking, loaded, pressure, heat-exposed, electrical, or safety-relevant part.
Apply only the fix supported by the pattern
- Low waves above long spans: shorten or support the span directly below the roof.
- Successive roof passes improve but never close: add one controlled total-thickness step.
- Exposure-linked pits and popping: use the exact maker-approved drying process and repeat the same slice.
- Soft glossy smears: test one supported heat, cooling, or top-speed change.
- Heavy ridges plus heavy dimensions: correct the proven extrusion or pressure issue in a controlled amount.
- Scratches or deposited PETG: clean safely, then correct why the surface or residue reached the nozzle.
Run one controlled PETG roof test
Use a small repeatable coupon with a broad flat roof over infill. Print two unchanged baselines, let them cool, photograph the face under the same light, and label support, top thickness, spool state, heat, fan, top speed, and flow. Choose one evidence-supported branch and print two more copies after changing only that variable.
Define a passing result before tuning
Accept a change only when repeated cooled samples improve closure and flatness without thin walls, weaker layers, heavier dimensions, corner lift, or new nozzle contact. Save the winning project. Return every failed branch to baseline before testing the next one.
Avoid these common failed fixes
- Turning on ironing first: ironing cannot build support under sagging lines or stabilize inconsistent extrusion.
- Raising infill everywhere: this wastes material when the issue is top thickness, spool condition, heat, or contact.
- Changing flow from the roof alone: a blind increase can create ridges; a blind decrease can create gaps and weak walls.
- Changing heat, fan, speed, and thickness together: a different result will not identify the working control.
- Assuming all roughness is moisture: require an exposure timeline and same-spool proof.
- Using one universal PETG recipe: formulation, nozzle, printer, profile, geometry, and cooling system change the useful range.
Use limits: when a smoother PETG roof still is not a passing part
A smoother final face proves only that this visible symptom improved. It does not prove the configured top shell is thick enough, the structure below it is sound, or the part will survive its actual job. Let the PETG cool, then apply the limit that matches the part.
- Decorative cover or noncritical face: minor texture can be acceptable after the roof is fully closed, no loose flakes or raised deposits remain, and two repeated prints are stable.
- Mating face, gasket land, sliding surface, bearing seat, or close-clearance assembly: reject ridges, pits, strings, or high spots that change fit, sealing, clearance, or motion. Check the cooled part against the actual mate; ironing is not dimensional proof.
- Loaded, impact, fatigue, pressure, hot, chemical, or outdoor part: top appearance cannot qualify internal roof support, layer bonding, voids, PETG creep, pressure integrity, chemical compatibility, or weather life. Verify the configured shell, inspect a representative section when appropriate, and repeat the real load, heat, exposure, seal, or cycle proof.
- Recurring nozzle scraping, dark deposits, smoke, a growing hotend blob, or material leaking above the heater block: stop surface tuning and follow the printer or hotend maker's safe inspection procedure. More top layers, flow, or ironing cannot repair contamination, nozzle contact, a loose hotend interface, or mechanical play.
- Food-contact, medical, lifting, pressure, electrical-safety, or other consequence-heavy use: a closed PETG roof coupon is not qualification for sanitation, dielectric safety, load capacity, or code compliance.
Confirm the intended solid-layer and minimum-thickness controls in the slicer's top-shell guidance. Use ironing only after the roof is already continuous and functionally acceptable; it is a finishing pass, not a repair for weak support, missing shell thickness, unstable extrusion, or nozzle contact.
Release the PETG part in this order
Let it cool, confirm the roof is closed, wipe-check for loose material, inspect the critical face for high spots and pits, verify fit and clearance, then run the representative load, heat, weather, seal, chemical, or cycle check required by the job. If any check fails, return to the matching diagnostic branch below instead of hiding the symptom with a surface-only setting.
Choose the next troubleshooting branch
If the roof is now sound but open-air hairs remain, use the PETG stringing branch. If the defect follows every extrusion region, use the under-extrusion branch. If the nozzle is scraping multiple features, use the nozzle-contact branch. For a polymer-independent roof failure, use the general rough-top guide.
Common questions
What should I change first for a rough PETG top surface?
Classify the cooled texture and inspect the sliced support below it. Broad sagging points to support; a roof that gradually improves but remains open points to thickness. Random pits plus an exposure timeline support a spool-condition test. Ridges and scratches require extrusion and contact evidence.
Should I add more top layers?
Only when layer preview shows reasonable support and successive top passes visibly improve. Add one controlled total-thickness step and reject it if it merely repeats pits, thin streaks, or raised ridges.
Can wet PETG make the top surface rough?
Yes, especially when the same spool worsened after exposure and roughness arrives with popping, bubbles, pits, fuzzy lines, or inconsistent gloss. Prove the branch with the same slice before and after the filament maker's approved drying process.
Should I lower PETG temperature or increase cooling?
Either can be a valid controlled test when the roof is glossy, soft, or smeared, but use the maker's supported range and change one variable. Reject any result that weakens bonding, increases warping, or creates poor extrusion.
Will ironing fix rough PETG?
Ironing can refine an already continuous roof. It is not a first fix for sagging over infill, insufficient thickness, unstable extrusion, moisture pits, heavy ridges, or nozzle drag.