Z-Axis issue, over extrusion, or warranty?

[Taco_Bob]

• Printer: P1S, default 0.4 nozzle preset, 1200 hours, 1.09.00.00 and 1.09.01.00 firmware

• Filament: Bambu PLA basic/matte, default presets for troubleshooting. Have printed with various abrasives and PETG in the past, but not recently.

• Plate: Textured PEI and CryoGrip Glacier

Been printing with my P1S for just over 6 months now. I’ve had my fair share of print issues and have been troubleshooting in stride, usually resolving them without too much frustration. Now however, I’ve hit a brick wall after coming back from a two week break.
Before the break, I had successfully printed my first HueForge. There was some color issues in the model however, so after the holidays settled down, I wanted to give it another go after addressing the model problem. Same slicer settings and filament as before, never left the AMS unit, but didn’t get far before noticing terrible distortions along the perimeter.

My initial thought was over-extrusion, but tried my go-to calibration model to work with something I’m more familiar with. Here the issue looks to be with bridging, which I’ve had previously during partial clogs. This time however I noticed the problem actually starts before the bridge, consistently on the right post. These posts are detached from the rest of the model, to test tolerance, and I noticed that while printing the post becomes detached from the plate and starts wobbling, ruining the rest of the print. If I secure the posts, either by attached to the rest of the model or cranking the temp bed with textured PEI plate, then the calibration model prints perfectly.

The prime line also seems to be occasionally detaching and getting caught up in the print.

To summarize, I’m seeing two symptoms that have appeared out of nowhere, after the machine set idle for 2 weeks.

1. HueForge print is a mess along the perimeter
2. Certain objects detaching from the plate, specifically thin lines and skinny objects with height. The bridging looks to be a consequence of this, not the primary concern.

So far I’ve considered if the problem is over-extrusion, bed adhesion, z-axis offset, or some combination. I have no clue how this has just started happening, without much of any real changes to trigger it. I’ve tried the following and I’m out of ideas at this point.

• Repeated cold pulls
• Swapped nozzles and hot end assemblies, now using a fresh unused 0.4mm stainless steel nozzle
• Switched plates (CryoGrip Glacier → Textured PEI), then tried again after cleaning
• Dried filament for two 6hr sessions at 45C; Desiccant and humidity levels in AMS look fine
• Filament calibration via BambuStudio, flow rate and flow dynamics
• Full printer calibration
• Routine maintenance; cleaning carbon rods, greasing z-axis screws, cleaning and clearing fans
• Tweaked retraction and z-hop settings, forgot to record specifics
• Replaced extruder assembly with new hardened steel version
• Tested various raised z-offset using startup g-code, up to 0.15mm
• Tested bed temp accuracy using contactless thermometer, all within a few degrees.There is a significant drop in temp at the bottom of the plate, where the prime line prints.

Updated with missing details and listed chronologically.

At this point, I’m considering looking into the warranty (printer is 6 months old), but would love to solve this without having to go through all that trouble. Any thoughts and suggestions would be appreciated!

[Olias]

Welcome to the community.

It may help to manage your expectations and adopt a more systematic approach to troubleshooting. Thanks for posting clear photos - they help.

Based on your post and the issues you raised, you are chasing too many variables at once. Rule #1 of troubleshooting anything: change only one variable at a time.

Let’s start with the last image. I’m not familiar with that specific model, but it appears to be one of the many “torture tests” that, frankly, add more noise than value. The one you’re showing is clever, but it offers little practical utility beyond the classic Autodesk FDM Test that has been widely used since around 2018. Even that test is arguably overkill, as it tries to evaluate too many things at once, takes a long time, and consumes unnecessary filament. You are better off using simpler, targeted calibration models.

The first step, purely for process of elimination, is to confirm the filament is dry. Do not fall into the trap of “it was factory sealed.” I’ll use my own example of “factory sealed” Bambu Silk PLA (now discontinued). Guess which spool was wet? PLA is often assumed not to require drying, but that assumption is frequently wrong.

CRITICAL STEP: Before drying, it is absolutely necessary to weigh the spool before and after. This is the only way to confirm that there was moisture present and then removed. Without this step, one has no proof that moisture played a roll. During drying, weigh it periodically over a 24 hour period, you will know when moisture has reached it’s lower limit when it no longer looses weight.

From the calibration model photo, there are visible issues that should be resolved before drawing conclusions about hardware faults or more exotic causes.

Once the filament is confirmed dry, manually calibrate it in the following order:

1. Temperature tower (some prefer to do this later; order preference varies)
2. Flow rate
3. Pressure advance - use the pattern method for quick diagnosis and the tower method (20-25 minutes) for higher accuracy
4. Maximum volumetric flow

Reference:

home

For video guidance, this channel does an excellent job walking through manual filament tuning using these calibration routines(he has 3 videos on tuning worth viewing):

Butter Pockets Prints

Only after drying and manual tuning are complete should you move on to optimizing wall thickness, layer height, print speed, or other secondary parameters.

[RetroSharky]

@Olias

Based on your post and the issues you raised, you are chasing too many variables at once. Rule #1 of troubleshooting anything: change only one variable at a time.

100% exactly that. Your comment sums it up perfectly. It makes no sense to test troubleshooting on a complex object if the first step hasn’t even been reached: Can the printer still print something simple like a cube?

That’s always the first thing I recommend when troubleshooting. Step by step.

@Taco_Bob Nevertheless, we must also acknowledge your efforts; it is uncommon to come across such a thorough error description, which is naturally extremely helpful. You did that really well.

As has already been mentioned, though, you are trying every possible solution and losing sight of the fundamentals in the process.

For example: You say that you dried your filament, but it may have deteriorated anyway, as Olias described above. Then you spend time unsuccessfully trying to find the error. That’s why you need to take it step by step. Yes, we all know it’s annoying, but you have to get through it. First, print a few cubes using the standard settings. Then you can move on to the calibration test.

[Taco_Bob]

I absolutely appreciate the advice! I certainly made some assumptions, based on my limited experience. I’m at a dead end, so if it means back to the basics, I’m all on board.

I’ll proceed with doing a test cube, though wanted to clarify a few things. Just filling in any gaps in my original explanation, before proceeding with your suggestions.

Each of my tests have only been with one change at a time, reverting to baseline whenever possible. I’m hesitant to believe the filament is the issue when I already dried, calibrated (flow dynamics+rate), and cleanly printed with every filament you see here just two weeks prior. This is where my concern lies; I ran into such drastic differences when little changed. That being said, one of the first things I did was dry everything for another 12hr at 45c.

Failed to mention in the original list, but I did re-calibrate flow rate + dynamic/PA with the purple, getting similar calibration results to the existing config. It might be worth mentioning that one of the flow rate lines did detach in a similar way as my prime line keeps doing. I also tried lower nozzle temps and reviewed cooling, as an attempt to improve the bridging, but never a temp tower.

With the test piece, I started by trying to address the bridging and overhang given that those are the most obvious and that is where I’m still at. What made me stop focusing on it as a bridging problem however, is when I watched it print and noticed the right post consistently lifting from the plate. The post then wobbles for the rest of the print, resulting in the mess you see up top. Should also mention it is always the same post, even when rotating the model at various degrees. If I affix the two parts in the model, everything prints MUCH cleaner, though still not perfect (see below image).

This to me suggests it is not strictly filament or extrusion related, but I accept that is me jumping to conclusions with limited knowledge.

[RetroSharky]

What I don’t understand is why you’re worrying about bridging at the moment - you can always do that later - the current status is that the printer is already having problems with a flat surface:

You own the P1, a high-performance race car, and you want to take it to the most extreme racetrack to identify any problems with the car. However, in your haste, you fail to notice that your tyres are flat.

In other words, you won’t climb the highest mountain if you stumble on the way there. What I mean is that your calibration tests are just wasting filament unnecessarily at this stage. They’ll be fine later on.

You shouldn’t worry about bridging at all for now - it’s obviously part of the problem, but playing around with the bridging settings isn’t the solution. The printer should be able to do that in the default settings. However, when I look at your pictures, any of them, I see much worse faults. In all of the pictures, several of them. Which means that something fundamental is wrong. So, for example, just pure examples, thoughts: Your extruder is defective - Your heating element is defective - etc. Just what level are we talking about here.

I’ll call a spade a spade: in your eagerness to get things done, you >>may<< have also ‘broken’ your printer calibration & settings and should reset everything to zero. This might not be the case, but it could be. Just a theory, could be completely wrong. Of course, in addition to the actual problem that something fundamental is wrong.

[Taco_Bob]

Seems I failed a bit in my clarification. I was only going through my thought process a bit and filling in any gaps.

Maybe this is me tripping at the starting line, but to me the issue was never specific to flat surfaces. The initial HueForge print that started showing issues, was mainly along the edges. The incomplete print maybe looks a tad rough elsewhere, but nothing like the perimeter. The rest I attributed to HueForge’s unique modeling aspect.

I’m not worried about bridging though. I only started with that because it was obvious, something I recognized, and (naively) thought it was something I’ve dealt with before; i.e. a partial clog. My focus shifted briefly to adhesion, then extruder or z-offset, based on the behavior I saw with the wobbling post. If adhesion behavior hasn’t changed, then maybe there is too much horizontal force, etc.

I realize my thought process on how I got here isn’t overly important. Again, was just looking to fill in details that could provide clues later on. I hear you, probably should have started with simple objects and gone from there. So here we are…

With a quick cube the prime line detached again and some slight over-extrusion(?) on top, but otherwise nothing too drastic to my eyes. Noticed some stringing and messy seams on the temp tower. My current temp of 220 is looking alright, though 215 seems slightly better. Went through flow dynamics (PA/) and flow rate again, no change, but did get another example of those lines detaching.

With this, I’m assuming it’s good enough to continue forward with different objects. First layer test, then overhang, then maybe something with denser infill? Any suggestions given the line detaching behavior? Or am I already making the same mistake and I should just continue on with the individual calibration tests?

I’m looking to avoid calibrating around the root issue, when nothing was previously wrong not too long ago (e.g. cranking heat bed temp just to secure adhesion).

[RetroSharky]

@Taco_Bob

Noticed some stringing and messy seams on the temp tower.

I’m just asking to be sure, just so we’re on the same page: You chose a temperature tower with G-code? I’m just asking because there are also some without G-code. So that’s what it looked like in your slicer? I’m asking because yours look similar, but it could also be due to the photos.

Otherwise, the printouts actually look quite good, so there’s nothing to complain about, but what really worries me is why your prime line has detached. You already mentioned that in your first comment, but now with the photo you can see what you mean - its slipping onto the printing plate.

Sorry for asking such a stupid question, I know it sounds silly, but I’m not standing in front of the printer, but could it be that you have a adhesion problem? Although that’s actually the wrong word - because you’ve already replaced the plate, it’s more likely that the filament is “not sticky”. That’s why the post, or whatever you called it, wobbled, because the Fialemt isn’t sticking properly.

So now I need to express this in a different way than “not sticky.” => “Layer adhesion problem”. However, it is not so bad that it ruins the print entirely, but it is noticeable that it pulls the line and also messes up your calibration part every time. That sounds like a partial clog - but that can hardly be the case if you’ve already replaced everything - or an error in the temperature. No wonder you’re hitting a brick wall. Now I understand the context.

One thing is certain: as long as it pulls your line around the surface, you probably don’t need to continue the next step. For some reason, the layer adhesion isn’t right, and as I said, I don’t mean the printer plate. That’s probably why the side surface of the cube looks so uneven and not smooth, see the lines too clearly. A little bit is okay, sure, but this is almost too much; it needs to blend together a tiny little bit.

Almost as if the filament were cooling down too quickly… as if the fan were constantly running at 100%… or cold wind were blowing into the room - yes, it’s a silly comparison, but I think you know what I mean.

I need to think about it, maybe someone else will come up with something. I read that other users had a similar problem, but unfortunately no one posted how they solved it.

Sudden poor layer adhesion

I just did the flow dynamics and flow rate calibration on my p1s. Everything was looking much better on my prints, until about 3 prints later. Suddenly it’s missing the layers, it’s ruining prints because of of extremely noticeable print lines. At some points it just doesn’t adhere and you can see through it. Other times the filament just starts fraying. It just slowly gets worse. First the lines are suddenly noticeable. Then, it’s as if it missed layers. Then it suddenly just doesn’t work at al…

[Taco_Bob]

No worries regarding any silly questions. I’d much rather it be something stupid I missed, then have to deal with whatever this is.

@RetroSharky

I’m just asking to be sure, just so we’re on the same page: You chose a temperature tower with G-code? I’m just asking because there are also some without G-code. So that’s what it looked like in your slicer? I’m asking because yours look similar, but it could also be due to the photos.

The temp tower I use does not come with the g-code set, but I make sure to set it myself.

It sounds like you’re starting to see something similar to what I was thinking. Adhesion is off somehow, but it doesn’t seem to be universal. Larger surfaces aren’t having any adhesion problems at the corners and still feel quite attached post-print, including the HueForge. Even then, a strictly adhesion issue wouldn’t explain the mess with the HueForge corners. I can also force better adhesion by cranking the temp bed up with the Textured PEI plate, which will help the calibration model print better* but not the HueForge.
*see earlier clean print in my 2nd post

I should also clarify, the print line is only getting dragged every so often; approximately 1/8th odds. That dragging is where my head is at, thinking the extra filament from over-extrusion is producing a bit of drag and given solid infill, then just builds up into a mess.

@RetroSharky

One thing is certain: as long as it pulls your line around the surface, you probably don’t need to continue the next step. For some reason, the layer adhesion isn’t right, and as I said, I don’t mean the printer plate. That’s probably why the side surface of the cube looks so uneven and not smooth, see the lines too clearly. A little bit is okay, sure, but this is almost too much; it needs to blend together a tiny little bit.

Almost as if the filament were cooling down too quickly… as if the fan were constantly running at 100%… or cold wind were blowing into the room - yes, it’s a silly comparison, but I think you know what I mean.

To confirm, you’re saying the layer lines look too defined here? Wouldn’t that also suggest over-extrusion? In regards to the cooling, the part fan is running 100% with Bambu’s default PLA preset. Except for the first layer at 0, which I usually increase to the first 4. Usually set aux fan to 0, but reset it back to the default 70 percent as soon as I was having issues. The room the printer is in is a good deal colder this time a year, but it’s still climate controlled and nowhere near the temps I assume people get away with in garages and shops.

[RetroSharky]

@Taco_Bob

To confirm, you’re saying the layer lines look too defined here?

Since you’re using a 0.4 nozzle, it should look more like this. I don’t have a cube right now, but I think you can clearly see the difference compared to your printout. That was just a quick, rough printout I made. You can still see the lines a little, but not as much.

However, on the A series, not the P series. Nevertheless, the quality should obviously be the same. Well, the layer height also influences this, but as I said, the photo only shows a quick, rough print in 0.16 mm. I don’t remember exactly.

When I look at your photo, it looks more like the filament lines are simply lying on top of each other. By that I mean it melted incorrectly, not enough. I could be wrong! - but that’s how the photo appears to me.