How can I improve this PETG print?

[GroovyAlpaca]

First off, this isn’t a failed print. It came out quite nicely, just not perfectly, and I’m curious to learn why and what I can improve moving forward.

This was my first time using PETG on my X1C. I had a roll of Blue PETG from my old SnapMaker printer that I had opened years ago, but never used. I figured it would have absorbed moisture, so I dried it at 55C for 5 hours in a Sunlu dry box. Would’ve dried for longer, but I wanted to start the print before going to bed.

I created a custom filament (feature available in Bambu Studio 1.8.0 public beta) using Overture PETG as the baseline. Along with:

• 0.1mm layer height
• Engineering plate
• Supports enabled
• 5 wall loops, enough that the part is solid throughout
• AMS + Flow Dynamics Calibration + Bed levelling

I left the door on the X1C open a crack during the print too.

The print came out almost perfectly, but there are some defects near the end of the print. This was a 3h20m print.

Anyone know what might have caused:

• The small defect near the top, looks like under-extrusion, or some conflict with the supports?
• The poor overhang surface quality, but only on one of the 45 degree overhangs
• The uneven color tone. The filament should be all the same color. Maybe just poor quality control?

Thanks for the help!

[sharcrypto]

calibrate your filament

[GroovyAlpaca]

I was hoping the flow dynamics calibration would help with that. Are you suggesting I use OrcaSlicer to run calibration tests?

[sharcrypto]

i dont use OrcaSlicer

[Olias]

The print does indeed look great. However, if you’re looking to get a “perfect print” the first option in your tool bag before going the route of dialing-in the perfect calibration, should be to run the print at 50% speed through the use of the “Silent mode”. That will at least rule out issues with the model or its orientation to the print plate. If it comes out perfect, then you know to chase settings values that are impacted by filament speed not temps per se.

This here looks like a case of under extrusion(1)

What is troubling is the waviness in your supports #2. But that could just be the oblique lighting. Either way, that would have me looking at chamber temps but you said you cracked open the door. Are you measuring temps? Again, nozzle movement speed can mitigate this to a degree but to really dial-in would require some experimentation with max flow rate, nozzle temp and travel speed.

I also am troubled by the void in #3. That too looks like an extrusion problem but it looks almost like the filament got snagged out of the AMS and the subsequent void was caused by an interruption in filament flow. Again, just a theory because the corresponding model opposite of the support structure looks flawless. So if a filament snag did occur, it was momentary enough not to affect the rest of the model.

[GroovyAlpaca]

Thanks for the help @Olias. I did a few more tests, recalibrated the filament, and printed at 50% speed.

Printing at 50% speed helped the issue marginally, but I’m still getting the issue on just one side of the model.
Additionally, I bought BambuLab Black PETG, dried it at 55C for 24 hours fresh out of the package in a Sunlu drier, and immediately tried printing just the problematic section at 50% speed right after drying. AMS, bed levelling, and dynamic flow calibration enabled.

Here are the results:

Left to right is:

• 0.2mm layer height, normal speed
• 0.1mm layer height, 50% speed
• 0.1mm layer height, normal speed
• 0.1mm layer height, 50% speed

Even with the BambuLab filament, I’m still getting undesirable artifacts, arguably even more so. And, with the Bambu stuff I’m getting buildup at the nozzle end - something Bambu specifically advertised was an improvement with their brand of PETG. Apparently not so :(.

I went back to Bambu Studio and noticed that the seam (white dots in image below) is along the right side of the model, close to where the artifacts are showing up on the print. Could it be that the filament is being retracted too much and/or extruded too little when incrementing layer height, thereby causing under extrusion when beginning the next layer line?

Are you familiar with this issue and how I may fix it?
Any help is much appreciated, thanks!

[Olias]

You’ve clearly gone through extensive measures to fine tune this model. Hat’s off to you for not giving up. I think you’re getting much closer to the perfect print.

I am familiar with the seams feature and in rare occasions using the seam painter I have been able to get different results but manually placing the seam “somewhere” else on the model. It’s worth a try.

However, I might also suggest some simple changes if you haven’t already done so. Attempt this individually otherwise your results will be hard to interpret.

1. Try orienting the model between 30 or 60 degrees off the Z rotation. Not 45 degrees as that is an angle that would make any issues with stepper motors ambigous. By moving between those two angles, if the problem moves then you may have issues with pullies needing slight lubrication(very slight like 1 drop) and or stepper motor tolerance issues for which I am not qualified to offer a remedy.
2. Alternatively, you can vary the angle of the infill. This may actually be the easier of the two. Again, if you see the problem move on the model, that is an indicator more likely to be related flow problems.

---

---

Alternative test I recently started experimenting with(under 2-3 minutes to print)

There is one uncommon test that has caught my attention that may reveal some hidden aspects of either your filament, build plate or flow rate. Uncommon because it fell out of favor when bed-leveling tech became the norm.

I found that first layer testing has been telling me a lot about both my filament flow and my bed hygiene. If my print bed had any contaminants, it showed very clearly. Likewise of it I used a new different plate and did not bed level it.

NOTE: This is a curiosity more than a useful calibration tool but it does reveal flow characteristics of different flow rate.

Here’s how to run this test:

1. Start with a cube primitive. Scale it to 50 x 50 x (2x of your layer height.) Note: if you go to low in layer height, the slicer will complain that there is no printable object and it will not print.
   

   

2. Under strength set these to zero.
   

3. Under Speed set first layer to equal top surface for an apples to apples comparison of layer performance once the first layer is laidd down.
   

Your model after slicing should look like this. Text is optional and I only use it to identify the test sample when doing repeat experiments. Note that there is only one string of filament on the plate.

Here is the results it produces and as you can see in this experiment, I slowed down the flow rate significantly to see if it would have an affect on some suspect filament I had(I eventually returned this filament to Amazon BTW because it couldn’t be tuned)

Notice the extreme rippling that occurred between flowrate of 0.70 to the normal 0.93. The 0.70 came out smooth BTW but produced gaps during solid prints.

---

---

A Long shot experimental feature in Orca.

There is a setting I recently discovered in Orca Slicer that supposedly will NOT do anything for Bambu. It was said to have been included to help out Ender printers. But it is intriguing and might be worth a shot. I couldn’t find a lot of data on this but I was using this on a buggy model where I was getting some weird cornering nodules. In my case, this didn’t help me but nevertheless it may be worth experimenting with. In my case, changing the angle of the model was the solution.

It can be found under advanced or developer mode settings turned-on, under the Global Speed tab at the very bottom. I could not find any wiki on this topic other than what you see in this help menu