Solution to printing Bambu PETG-CF with X1C (.4 and .6 Nozzle)

[Damocles]

Right off the bat, please note: I have researched all available post, threads, videos, and comments found relating Bambu PETG-CF and the X1C. Some are not included in here as they either are redundant (same user posting the same info) or they do not provide sufficient evidence for the settings

Please read what has already been done before commenting. This is long and comprehensive, but should provide a single resource to what is available on this topic

Like many of you, I have been plagued with the desire to use Bambu PETG-CF and yet, been met with the inability to have acceptable results. I hope to solve that here by providing what I have found, while also seeing if we can finally consolidate a definitive solution to this. I have titled this topic as such so that once a solution is nailed down, I can update this first post.

The purpose of this topic only applies to Bambu Lab Filament and the Hardened .4mm and .6mm Nozzles. If I don’t set that limit now, we will have too many other variables. Yes, other companies make good and maybe even better PETG-CF, but that does not solve the problem at hand. Feel free to mention them as a point of comparison, like “Here are my settings for Bambu but with ___ filament, I did not have these issues and used these settings”. This could help us determine what the underlying issue at hand is.

The predominate issues many encounter seems to be:

• Blobs on the print - Small dots of defects either at the start/end of a layer or randomly in the middle

• Tearing of top surface - Rough or completely ripped top layers that can expose the infill or just be cosmetic. Can also accumulate on the nozzle.

• Nozzle collecting filament (not always the same as the blobs) - though related to blobs, this typically is much larger amount of filament that collects on the nozzle and will either randomly be deposited on the print, or can be manually removed/falls off. Either way, this results in areas lacking filament and mimic under extrusion only to be followed by a burst of filament (clog?)

• Poor Bridging and overhangs - Strong warping or outright failures on short bridges and strands hanging down on overhangs. (potentially a byproduct of these others issues so I will not be diving too deep into this)

• Failures on long prints - Some combination of one or more of these compounded, by long print times.

There are quite a large number of complaints/request for help when it comes to this filament on this forum and other outlets. Unfortunately, none seem to fully solve the issue. Here, I will compile the some of the solutions/suggestions I have found. These are not really in a particular order, besides first 3.

1. Dry your filament - For some reason, this is a highly controversial and divisive topic. I am guessing it is because in the early bedslinger days, we rarely dried our filament. We also only had PLA and this is not a bedslinger. General recommendations seem to be 65-70 °C for 8-12h. YMMV

2. Calibrate your filament - PETG-CF seems to be hit or miss with the autocalibration. I have found that it can overestimate the values but I have also had several good calibrations. Calibration is needed because if the printer puts out even a small amount of extra filament, it will aggressively collect on the nozzle.
   In general:

• Flow Ratio should be around .94-.96
• K Factor of 0.024
• Volumetric flow rate of 12-18 mm³/sec

For reference, if someone has properly calibrated their filament, I will be referring to it as “Fully Tuned”

3. Create proper print settings - This one is going to have a bit of back and forth as some people have found great success with the stock profiles while others have had headaches. I believe this where many of the problems and solutions are.

I will be assuming the top glass is off for all of the prints as that is what Bambu recommends.

Print Settings:

I am not sure the best way for this to be formatted at this stage as we are still compiling the results.

Right now, I think melding the predominate issues (with links to their existing topics) and providing their listed solutions/what I have tried could be best. My thought is that if I link to the original post, but compile the solutions here, we can see what worked and what did not.

As for how to handle the nozzle issues, I am not sure. At some point each will get their own section (or if it is determined the .4mm is not advised then just the .6mm will have one). For now, both will exist here so please specify what nozzle you are using

Blobs on the Print -

User @ant posted Help Tuning Bambu PETG-CF earlier this year and @Balint posted Bambu PETG-CF sticking to nozzle, unable to print
Their issues seem similar and many solutions were tried. For ant, the original settings were using a .4mm Nozzle, the filament was dried and fully tuned. .2mm layer Height, Nozzle 255C, and Gyroid Infill at 15%. They did have the lid and door closed at the start. @3DTech then provided good insight into why they recommended what they recommend.

Their suggestions were:
- Min layer height of .2 with a .4mm nozzle and .18 with a .6mm nozzle
- Bridge Speed of 30 mm/s
- Cooling Min/Max of 40% - 20s and 90% - 6s respectively.
- Disable Slow Printing down for better later cooling
- 25% overhang threshold with 90% Fan speed.

This was then followed up with ant saying new testing using these settings found success (see his post for credit to other users)

• Rectilinear infill. Overhang Speed Reduction of 0-20-20-10. Reduced Chamber fan speed (no value given). Seam Rear. This was using stock PETG-CF profile for the .4mm nozzle.

For Balint, they seemingly had tried most of these already with little to no success. The filament was not specifically calibrated as they stated they found no difference. Below is their settings and I am confused as to why all failed.

• switching nozzles: 0.4mm and 0.6mm
• switching layer heights: 0.2mm nozzle, 0.24mm and 0.3mm with 0.6mm nozzle
• the nozzles are not clogged, damaged and very clean when starting the print
• coating the nozzle with PTFE lubricant
• wall order: outer walls first
• wall speeds: 60mm/s outer, 80mm/s inner, 100mm/s infill
• infill type: gyroid
• infill overlap: -2%
• chamber fan set to 100%
• part cooling fan set to 10-30%
• everything else stock Bambu PETG-CF profile
• everything else my print profile based on these Bambu print profiles:
  0.3mm Strength (0.6mm nozzle)
  0.24mm Strength (0.6mm nozzle)
  0.2mm Strength (0.2mm nozzle)

Tearing of top surface-

@kragghc posted PETG Tearing Top Surface where the filament was dried for 12 hours, calibrated, and tested using various settings on a P1P with .4 Hardened Nozzle. I will say, their listed flow ratio of 1.03 and PA of .032 seems to be too high IMO. Nozzle 265C and Max Vol Speed of 13.

Cooling is similar to what 3DTech had recommended in the other post with the addition of Retraction 2mm and Z Hop of .4mm (the Z Hop should be useless as the default is already .4mm. Retractation default is 1.4mm)

There was also a claim that sparse infill was being printed lower than the main print layer. This is the only time I have seen this mentioned, but could be worth exploring before just dismissing.

This is the first time that major changes to the print settings were done. Here are the ones that I believe would actually effect the print:

• Top/Bottom Shell - 5, Bottom Surface - Archimedean Chords, Internal Solid infill - Mononic Line, Disable Ensure vertical shell thickness, Enable Infill Combination, Enable Avoid Crossing Walls, Gyroid 50%, Wall Loops - 4, Support Style - Snug, Bottom Z Distance .35 (.2 Default), Top Interface Spacing .1 (.5 Default),

• For speeds(default value in parenthesis): Gap 200 (250), Inner Wall 200 (300), Internal Solid Infill 200 (250),Outer wall 140 (200), Sparse infill 200 (270), Top Surface 150 (200).

It later became suggested that the issue was due to partial clogging of the nozzle due to the Aux Fan. This never fully proven, but is now the 2nd time that has been suggested. The failure also occurred on a sparse infill layer.

@wiremeup came in to suggest settings that are similar to what we saw on the Blob section, but with the changes/additions of: Grid infill and patterns that avoid sharp movements like Concentric.

Retraction should be “high” (which later was suggested to be 4-5mm) and fast, with flow ratio “low”. I disagree with the Grid infill as my testing shows it to be harming in all cases and all filaments. And others did as well. Just adding it for posterity. They also later stated you must use the .6mm Nozzle.

They were really institant on Grid Infill and did back their claim with many photos of very nice parts, but it was all printed on another printer/filament.

kragghc later stated they tested Rectilinear but found no benefit. They also tried 50% print speed, 2-6 walls, 25-50% infill, lower Vol Speed to 10, and using the Generic PETG profile but still failed. Not sure if these were at one time or not.

No real solutions were had, just these suggestions.

Nozzle collecting filament

We have now reached the existing megathread by @ohrenstoepsel Mega Thread

This is currently at 106 post, so any summery is next to impossible. It also makes finding solutions time consuming.

Many of the suggested solution are similar to what we have seen above, but here are the major ones:

.4mm Nozzle

@djeZo888 suggested Nozzle 265C (which was later dropped to 255C), Bed 80, PA of .032. Reduced speeds the same as in the Top Surface with the addition of Support Interface 60 (150), Support 100 (150). Print with door and lid closed. (the example print provided is the BentoBox which is small and has very little infill)

3DTech is back again. They suggested the Slice PRP paint but did also state they do not use Bambu brand PETG-CF. They also added changing Retraction settings to:Length = 1.1 (1.4), Z hop = zero (.4), Retraction speed = 40 (30), Deretraction speed = 30 (30), Travel distance = 0.6 (Not sure if this is the Travel Distance Threshold or the Wipe distance?).

We are also back with 20% infill and Rectilinear/Gyroid suggested.

@s.3 makes a lot of contributions here and does a large amount of testing. The main issue is that they are using a cht bimetal high flow nozzle, so a variable to keep in mind.

They did suggest two things strange early on: slightly pulling down the silicon sock and Printing the first layer at 200C Bed 90C. This did seem to work for them and they are the first major long test of a blob free 7 hour print.

Balint also provides some new insights after their original post. Again, drying and tuning are stressed and using Rectangular (Maybe meant Rectilinear?) infill,

What is new is: Reduce Infill Overlap, Disable Vertical Wall Thickness, Match Inner/Outer wall speeds, change wall order to Outer/Inner and Inner/Out to compare, Enable Arachne,

.6mm Nozzle

wiremeup This is now were we finally get more suggestions with the .6mm Nozzle. I am going to use his photos for these because there are so many changes. These are quite the departure from what has been previously suggested.

ohrenstoepsel later follows up with some more .6mm changes suggesting:

• 260C, Gyroid 15-20%, Aux Fan 20%, Chamber Fan 70%. The rest is stock profile.

@r00t did more testing and stated they found disabling Z Hop on layer change and before travel (not sure where these settings are), increasing retraction distance and speed, and raising the temps to 270C help them the most on their P1P. They decreased Flow Ratio to .89 and decreased line width (not value provided).

@kungpaoshizi suggested a gcode addition ( G29.1 Z0.03 ) that is also mentioned as a solution to the top surface/long print quality. Now other threads comment on where this should be made, but the general suggestion seems to be at the very end of the Filament Start Gcode.

The final major addition is by XxBonnyxX (I ran out of allowed tags) who used principles from this Bambu Wiki Page for PA-CF on the PETG-CF to seemingly good success. Wiki Page for PA-CF. I am not sure what exact here they did besides drying and printing at a max of 100 mm/s.

Again, all of these have been suggested but no real consensus has been made.

What’s Next? and TLDR

This is where you come in. Hopefully since all the OP’s and those who contributed to all of these post have been tagged, we can get everyone in one topic and hopefully stay on topic to get this solved.

• While there are a few post I have not mentioned here, the major post have been. There are also post on Reddit but they do not seem to give much detail or provide the same input seen here. Same with the Facebook group besides one on their mods who always has great PETG-CF prints.

  • This is what I have tested. However, after compiling this I think I need to make changes.:
    • I have tried many of these settings to limited success. Drying and Calibration are assumed, Nozzle seems to be either .4mm or .6mm, but I am using the .6mm.
    • Gcode additions have not decreased blobs for me
    • Lowered Temp to 250C (First layer 240C)
    • Aux fan 0, Chamber 70. Part Cooling 0-20,75-15. Slow Printing Down enabled, Overhang threshold and speed 10% and 90%
    • Retraction set to 2mm, Z Hop .4 (it seems I should disable this), Enabled Retract on Layer and Wipe while Retracting, Retract amount before wipe 85%.
    • Layer Height of .24, Classic Wall, Inner/Outer/Infill/, Avoid Crossing wall 400%.
    • 2 walls, Cubic Infill 20% (doesn’t seem to help), disable Infill Combi, Enabled Ensure Vertical Shell
    • First layer 35, infill 55
    • Outer wall 80, Inner 120, Sparse 100, Internal Solid Infill 120, Top Surface 100, Slow Down for Overhangs 0-25-15-10, Bridge 20, Gap Infill 250.

TLDR

• The main consensus seems to be as follows

  • dry filament and calibrate (potentially drop flow ratio to .89?)

  • Change retraction distance and speed on .6mm nozzle (anywhere from default to 4-5mm, Speed of 40mm/s all the up to “high” which was never given a value)

  • Edit cooling values of both the part cooling (Min/Max of 40% - 20s and 90% - 6s), Aux Fan (anywhere from 0-30% ), Chamber fan (suggested anywhere from 30% to 100%

  • Add G29.1 Z0.03 gcode (not proven)

  • Disable Z Hop?

  • Drop all print speeds potentially down to 100mm/s with very high bridge speed tested in one post.

  • Rectilinear infill is repeatedly mentioned. I really think this is where the problem stems from

These are the settings that repeatedly came up. It does not seem to be that temp is playing a massive role here as those have been changed with the issue persisting. It mainly seems to be that the infill is causing this, but I am not sure what the solutions is.

I believe we need to run a orthogonal (Taguchi) array to isolate variables here. With a standardized part (that takes sufficiently long to accumulate an error), nozzle size, and settings, we could all run a single test but compare them via a point system to see what changes things and while still allowing for two variable changes at a time.

Thanks for reading and I hope this wasn’t too boring.

[Damocles]

saving this spot for the day a solution is found

[RMB]

Man, that was a lot of reading. Are people really having that many problems printing with Bambu PETG-CF? I’ve not used it, only generic no-name and prusament which I’ve had no problems printing with. The only issue I’ve ever had with it (and it affects all PETG filaments I’ve tried on this printer) are poor overhang performance. I realised that this printer needs slowing down a lot on some prints so that each layer has enough time to cool before the next layer starts. If it hasn’t sufficiently cooled it will droop down with the force of the next layer on top of it, making the print look ugly. This normally only happens on overhangs because only part of the layer line is on top of the previous layer, the rest is in mid-air and why thinner layer heights work well to minimise this. But it’s not practical to use less than 0.16 with CF due to the infused fibres.

[Damocles]

I appreciate you reading it. The format is a bit janky, but it was the best I found after trying a few options.

As for the PETG-CF, I would absolutely say that a large amount of people have issues with it. However, I think there is a large amount who do not due to their print geometry and time. It seems that shapes with limited infill, short, and lower print time find more success than longer ones.

A prime example of this is the BentoBox, which was printed flawlessly out of PETG-CF on a .4mm Nozzle. But being mainly walls, it is also not very taxing.

During this testing, I have been printing parts of the BLV AMS Riser in PETG-CF. These are 15+ hour prints. The first two attempts were poor. I did print a new section to much greater success, and would place the print quality at around 50% (100% being mostly defect free print). This was much better than the prints that outright failed.

I agree with on overhangs, to a point. This print actually has a decent bit of overhangs and here is what I found:

For overhangs, you need the line currently printed to have two things: Bond to the inner layer that was printed before it, and support from the layer below it.

The problem is that these can be hard to balance. Higher heat will increase the bond between the inner and outer layer, but cause deformation of the previous layer supporting it leading to strong but droopy overhangs.

Or, you can increasing cooling where the outer wall isn’t dumping heat into the previous layer and will not droop, but will have less adhesion with the inner wall and the previous layer. Technically, you can also alter speed but this can cause a double whammy of dumping heat into both layers while also not having enough cooling. CNC Kitchen has some good videos on why sometimes, fast overhangs and bridges can be better as you are getting the hot nozzle away from the layer to cool.

Conversely, fiber filaments typically have better overhangs due to the fibers decreasing how the material deforms.

The problem with overhangs/bridging on the PETG-CF seems to be multifactorial. PETG already can be very “liquidy” and likes to ooze, so bridges and overhangs allow that to be exacerbated. Add in that PETG also loves to stick to itself and hardened still and now you have buildup on the nozzle. These just add on to all the aforementioned issues with PETG-CF.

[SimEyeSee]

Have you come across this article? It is the most comprehensive one I have found about PETG.

It might have a suggestion you haven’t tried yet.

PETG Filament - Complete 3D Printing Guide [Updated for 2022] — STANDARD...

[ohrenstoepsel]

Problem is not PETG the focus is on PETG-CF

[LagoaLabs]

your response is fair but since it’s only a max of 15% carbon fiber and the material still being mainly PETG I think we shouldn’t dismiss any tips for PET-G that quickly :).

Anyway anything that is PETG always has had the same issues in general with clinging to the nozzle. This is why E3D came out with their socks (specifically for PET-G) and even then it’s just recommended to put some PTFE on your nozzle if you don’t want anything to cling to it.

@damocles; I’ll go through everything you wrote and do some tests based on the best information that you have collected to see if there is anything that I can still add to this discussion that might result in better prints. Did you already come up with a design that you find suitable for apples to apples comparisons? Would the bambu cube scaled up not suffice or do you think it’s too short of a print?

[kungpaoshizi]

Nice thread but I dispute the ‘not proven’ on the G29.1 Z0.03
PETG has had issues sticking to the nozzle since the beginning of time. And raising the Z offset works every time. I spent hundreds of hours trying to find a solution to PETG epic ball fails, random torn surfaces, etc. It was a random comment long ago by someone I wish I could credit.
Carbon fiber additions are just coincidental

[wiremeup]

I’d like to link to a post I made on a previous thread. I believe I have discovered a bug perhaps in the slice or maybe its just poor choices in general. Gyroid + infill combination = incorrect infill gcode.

PETG-CF Solid Infill/Top layer tearing on larger prints

I just noticed something in my slicer. If you choose Gyroid and ALSO infill combination it does not (in my view at least) code the infill properly. Gyroid is supposed to be completely line in top of line and slowly twist around back and forth. Bascially there should be no gaps between layer lines. If you choose infill combination it creates huge gaps between the layers. I believe this is a bug and probably has caused my prints to fail in the past. Don’t know if this is your issue but its definit…

[mmajchrowicz75]

Ok in my situation automatic arrange with object rotation helps a lot. With that is not possible print some separate details.

[Damocles]

I will respond to @SimEyeSee here as well since these are connected. While I have not see that article before, I read over it and many of the solutions are similar to what I posted. The main reoccurring one being a Z- Offset.

To @LagoaLabs, you are correct. While the CF content of Bambu’s filament can apparently vary a decent amount, the underlying principles of the base filament still applies. ASA-CF settings wouldn’t work for PETG-CF, by maybe some of the changes incurred by adding to CF to any filament would apply here?

The PTFE on the nozzle seems to be quite hit or miss and I have not found any clear winners in that arena, but like you said, testing may need to be done. I have ran several test myself on practical prints which some success, but I still do not have a clear test model. From what I am experiencing, it seems you need:

• Sufficient size for errors in infill to occur/accumate (small models just don’t work)

• Some need for supports to be used

• Long print time for clogs or nozzle build up to occur.

I don’t think the cube would last long enough, but maybe the calibration model with some supports would be? I’m open to all suggestions.

[Damocles]

I should have worded that better. By “not proven”, I meant that is not the singular cause and solution to the issue. I actually found going up to 0.04 to be a pretty strong improvement. That, coupled with 20% Rectilinear infill, concentric top/bottom layers, and speed changes to allow for a decent 15 hour print. Still not something that I would say is good, but passable for something where looks do not matter.

[Damocles]

holy moly! You might be onto to something here! It seems that rectilinear might also not do this as I cannot find anywhere it occurs. That, or it might not be as pronounced.

While searching the GCode, I have also found that if interface layers on support are used a new issue can occur. Say you have a layer height of .24mm with a bottom interface of .2mm.

In my model, my first interface is at layer 761 at a height ot 120.24mm. The next layer (762) in the Gcode has a height of 120.3. 763 is 120.54 and is not until layer 764 that the 2nd interface layer printed at 120.65. This pattern continues.

This creates plenty of opportunities for filament to sag or buildup as gap between interface to interface layer is .4mm! It seems that the top and bottom Z distance applies to each interface layer? Not sure if that is intended or not.

This, coupled with what you posted could be contributing heavily to our woes.

[user_793764204]

Following. Tons of good troubleshooting here.
I’ve had a few beautiful prints with the stock settings (a couple 2 hour, and one 5 hr). Then I started printing another 5hr print and I noticed a couple hours into it, it wasn’t printing. The nozzle got clogged inside, and there was more than normal build up on the outside of the nozzle. It was a pain getting it all out, but I’m printing again today. We’ll see how it goes.

[wiremeup]

Precisely. I have tried this in software and the same anomaly occurs in Orca, BambuStudio and SuperSlicr. There is just too much gap between IC layers. Not only do the lines sag because of this but the points where they cross tend to not melt together but rather create speed bumps for the next layer to smash through. I printed an object the other day that left pits and bumps in the very top layer that I can only attribute to these “speed bumps” left behind from the crossing of Infill combination layer lines. I have reverted back to gyroid with no infill combination and I have had no failed prints since then. This type of problem obviously can become compounded exponentially over time the longer the print is. Here are some pics- 1 is my ratrig doing proper gyroid with no IC. The next is the same part and printer but with Grid infill and IC turned on. Then the last is wait for it…A brand new P1S using same filament (but its a little wet) with grid/IC turned on. For some reason it actually looks normal. Not sure if its because the filament is wet but that is another variable I’m throwing in here.

[kungpaoshizi]

Good to hear on the .04. With my old cartesians it was more in the range of .3mm, so the repeatable accuracy of this printer is very welcome.

As far as where looks do not matter, idk what you’re talking about
The following was without supports, OOBE default everything else for PETG other than adding the G29.1 Z0.03 to end of filament startup code, one of the finest prints I’ve seen.

[boogle]

What a great idea! I wanted to print top vents for my P1S in Bambu PETG-CF and it started out as a pretty cursed filament. With the help of content in this thread and others I managed to improve the quality drastically:

To do this I used the settings below. Critically once the settings are in place run a full calibration on the filament. Both flow dynamics and flow rate. My K value barely changed at 0.017, but the flow ratio was way off in the stock profile. I’ve ended up with a flow ratio of 1.0707. I think the built-in profile is designed specifically for the X1 with the idea that automatic calibration will set these values correctly.

I’ve also tweaked the temperature down to 255 from 250. 260 would work as well. Based on the temperature tower I printed, 255 was a halfway house between 250 and 260 in terms of advantages and disadvantages. I ended up preferring 250.

For me I had to print with glass top off and front door open.

Anyway, on to the settings:

Filament

Temperature: 250
Flow Ratio: 1.0707 (calculate your own)

Filament Cooling

Min fan speed: 40%, 30 layer time
Max fan speed: 90%, 10 layer time
Keep fan always on
Slow printing down
Force cooling for overhangs & bridges
Overhang threshold: 10%
Fan speed for overhangs: 100%
Aux fan: 30%
Exhaust fan: 70%

Filament Advanced

Add to the bottom of start G-code:

; Change Z offset
G29.1 Z0.03

Quality

Precise wall (experimental)
Avoid crossing walls
Max detour length: 200%

Strength

Top / bottom shell layers: 6
Infill: Gyroid (shouldn’t matter)

Speed

Outer wall: 100
Inner wall: 150
Sparse infill: 150
Internal solid infill: 150
Top surface: 100
Gap infill: 150
Bridge: 50 external, 150% internal
Normal acceleration: 5000
First layer (acceleration): 1500

I also used the Z offset gcode.

I’m just printing the actual parts now, we’ll see if they turn out OK or not. I’m sure there can be more tweaks made, but I need to get this stuff printed!

[wiremeup]

I can confirm now on several printers and both PETG and PETG-CF that machine g-code z offset set to 0.05mm has resolved 98% of the issues associated with this stuff. ZERO filament collecting on the nozzle now and ZERO blobbing even after several hours. I am also using 30% fan and 75% on forced overhangs. Speeds slow down gradually from 75mm to 15mm for all the bridging. I use special settings in the filament cooling to force 30% fan for everything but overhangs. Since I am printing production parts to be sold I need the surface finish to be perfect and consistent on every layer.

[wiremeup]

I also out of necessity for my particular parts had to adjust the line widths to get my overhangs to stick to each other. Specifically inner and outer walls. The sweet spot has been (0.6mm nozzle) to use 0.62 for everything but top surface where I use 0.60. Then outer walls 0.78 and inner walls 0.56. These settings force the outer wall lines to always be no less than 40% over the top of the previous layer’s outer wall. This is specifically to address issues with chamfers and filets on the bottom 3 layers or any other layers where there is a filet rising outward from the center…

[Landin]

Hello,
I’m a little out of place here because I haven’t used Bambu-labs PETG-CF, however I bought a PETG-CF from Eryone and was having similar problems. Weirdly enough the solution ended up being printing at 210 degrees. It didn’t even occur to me at first because I didn’t go below 230 with my temperature tower. Would be curious to know if this helps with Bambu lab’s PETG-CF as well.