X1C PETG Issues (Multiple Brands)

Finally decided to bite the bullet and try some PETG printing. My goal was to print that popular riser for the top of the X1C.

I bought a few rolls of PETG from Amazon from a cheap but well rated brand. I was able to print a couple of smaller parts successfully but once I moved to the larger bits I started getting really bad, I guess, adhesion issues? It’s like the filament isn’t quite sticking to the bed and then the nozzle is pulling it back up again. The below images, the first 3 are the Tinmorry brand I got from Amazon and the 4th is Bambu Lab PETG Basic using the built in profile.

Both brands have issues. What I’ve tried with the Tinmorry brand is:

  • Dried filament for 12 hours in Sunlu S4 on PETG setting
  • Washed both sides of textured PEI plate
  • Tried printing on both sides of textured PEI plate
  • Used needle on nozzle at 250 degrees in case there was a blockage (as recommended by the Wiki)
  • Ran fresh calibration on the X1C
  • Ran through the various calibration prints on Orca Slicer using this guide
  • Tried printing with settings on spool (Nozzle 220-250, Bed 75-90, no cooling on first 3 layers, 50/90% on other layers)
  • Tried printing with settings found on this forum
  • Tried built in Generic PETG profile
  • Lowered Max Volumetric to 10mm/s

The Bambu Lab filament seemed to print a little better but still seems to be tearing itself off the plate.

I’m at a loss as to what to try next. I even printed something in PLA in amongst all the testing to make sure it worked OK and it printed just fine.

Hopefully I’ve missed something simple and you can guide me in the right direction. Thanks in advance

Some additional info if it helps. Printer is X1 Carbon with AMS. It’s on the Anti-Vibration feet and using the Bambu Lab 4-1 PTFE adapter. Hardened Steel 0.4mm Nozzle. Prints were done with the door open

Edit: These are the tweaked settings I came to after following Orca calibration and reading some Forum and Reddit posts. But these still cause the issues in the images above

More than likely too fast. Basic PETG speeds under 100mm/s for reasonable quality. The first set of pics I’m assuming are the non-Bambu filaments, they’d probably string more, the Bambu PETG tends to perform better. No Aux fan as it’ll warp its head off with that big flat surface, door closed and lid on and preheat the chamber a little to get an even first few layers.

PS. Plate should be spotless, sink of hot water with Dawn (just in case you haven’t)

1 Like

I’ll wager that this is classic dirty play syndrome.

The simplest way to rule out a dirty plate is to do a first layer test. The easiest way to do that is to create a cube primitive that has the dimensions of 240x250x0.2mm

Here’s a post where that issues is discussed lower down in the thead on that issue with examples.

Alternatively, you may have to tune your filament but my money is on it being a dirty plate. If you state that you cleaned it with IPA, that isn’t good enough. You will want to use dish soap and run it under very hot water until you can visually see that all the soap has washed away. You should not need a sponge if you’re doing it correctly. The dish soap if smeared on correctly will dissolve any oils or contaminants off the plate and it will be obvious because the soap residue is easy to spot.

2 Likes

Interesting. I’ve actually only cleaned the plate with dish soap and a sponge, I didn’t think IPA was advised on the textured plate.

I did just print a pretty successful benchy (after just washing and drying the plate again) with minimal stringing so maybe you’re right.

I’ve got a first layer test started just now so that should rule in or out dirty plate

I’m not sure where you heard that. IPA is a universally safe solvent for everything we use in 3D printing with of course PVB for which IPA is a solvent.

You may be thinking of Acetone which will attack some of the build plates such as the engineering plate and high temp plate. And how do I know this? I found out the hard way.

Here’s a test I did to my engineering plate. You can see where the Acetone started to dissolve the coating. But if you’re ever in doubt about any solvent, do as I did which was to take a cotton swab and apply a generous amount of the solvent and then rub vigorously. It should be obvious if the underlying coating is vulnerable to whatever substance you’re testing.

1 Like

I guess this test suggests it’s just a dirty plate (and I suppose I’m not as careful handling the plate as I thought)

1 Like

Nobody ever is. :yum: But you can see that the first layer test leaves no room for ambiguity. It is the white glove test of 3D printing.

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.