I’m currently trying to print this lamp in vase mode with my new P1S.
The last two prints I did, the print head paused during the print and messed up the entire thing. We can clearly see it stopping in the video.
https://www.youtube.com/watch?v=XmpOmTg_tb8
The problem happened at different heights for the two prints, so I assume it’s not caused by a faulty G-code. I already successfully printed this model once a few days ago.
Any idea what’s causing this behaviour?
try changing the sd card. also, make sure all the software is up to date
Thanks for posting such a clear video - it really helps take the guesswork out.
The issue is pretty obvious from the start. At the 0:06 mark in vase mode you can see layer separation beginning, and the wobble of the model is definitely contributing (though not the only cause). What’s happening is that with each pass, the nozzle is laying filament slightly off from the previous layer because the print is shifting ever so slightly. Even though you had one success, this is a tricky issue to solve consistently.
Here are some things to try:
Slow down the print - Either from the speed menu or directly on the printer. Try “Silent Mode,” which cuts print movement speed by ~50% (though it doubles print time).
Tune cooling and temperature - Experiment with fan speeds (in the filament profile) and nozzle temperature. A slightly higher temperature can help with adhesion, while stronger chamber cooling can stiffen layers more quickly so the next pass has a firmer foundation.
Check your filament - You didn’t mention what material you’re using (PLA, PETG, etc.), but the photo shows under-extrusion lines, which suggests the filament is out of calibration.
One good way to do this is with a Max Flow Rate (aka Max Volumetric Speed) test. It prints a vase-mode tower at increasing speeds until gaps appear - exactly like in your video. You then measure where the gaps start and enter that value into your filament profile.
Here’s the problem - the best version of this test is in Orca Slicer, but with the latest Bambu firmware you’re locked out of Orca unless you enable developer mode and run LAN-only, which most new users don’t. If that’s the case, your fallback is to look for a Max Flow Rate test on MakerWorld. It won’t be as polished, but it’s better than guessing. Since I don’t use MakerWorld, maybe another member here can suggest a reliable one. Be careful though: most “test models” there are just fake uploads with no G-code, clearly posted just to farm points - yet another reason to be cautious with MakerWorld. You’ll want to check the flow box in the slice preview dropdown and make sure it shows the gradient flow pattern like below.
Here’s an example of the test:
And here’s a step-by-step tutorial:
Like most things in 3D printing, your mileage may vary - you’ll likely need to try a few of these steps in combination. Even small changes (like leaving the door or lid open) can make a difference and might explain why it worked once but not again. Keep experimenting and you’ll get it dialed in.
Thanks a lot for your very detailed answer.
I do indeed have two issues with this print.
The first issue is the one I described in my initial post: the print head stopping for no reason. For that, I’ll try to replace the micro SD card with a new one.
The second issue is exactly what you described: The model is wobbling during the print. What I’m observing is that the first layers seem to be okay until a certain point where the extruded PLA is “rising” above the print level. After that, it gets worse and worse and the nozzle pushes the printed model downward, causing it to wobble until the print line breaks. With some luck, the line eventually catches up to the model and gets back on it. That’s the breakage you can see in the picture below.
To solve the problem, the print speed has been reduced significantly (see the numbers below). I don’t know if I can decrease it more or if it will bring any improvement.
Regarding the cooling, the bed plate temperature has been reduced to 30°C without any adhesion issues. I tried running the aux fan at 100% speed and haven’t had better results, maybe even slightly worse.
Another interesting detail is that the breakage always appears at the same spot on the model (bottom left side of the print plate). I printed an air flow deflector to homogenize the cooling flow but saw no clear improvement.
I use Bambu Lab PLA Matte filament that I dried and stored in the AMS at 10% relative humidity. The print is made on a textured PEI plate with a 0.4mm nozzle. I have never calibrated the max flow rate.
I’ll have a look at the test prints you sent and see if I can find better settings for the print. If you have any remarks regarding my settings or printing method, please let me know.
A few misconceptions worth clearing up. Your build plate temps are fine - if you weren’t getting first layer adhesion, that would be the first suspect, but you are, so that’s not the issue. One troubleshooting step eliminated. 
The airflow deflector feels like “a good idea” without proof. I’ve gone down enough of those rabbit holes to know most mods sound better than they perform. Unless verified, they’re usually clever solutions in search of a problem. In your case, airflow is more likely helping than hurting. What we want is part stiffness, which vase mode reduces, so more airflow is generally better.
Note that matte filament is notoriously fickle in vase mode, much like silk. It just doesn’t like to bond to itself unless you bump the temps toward the upper limit, around 250 °C.
If airflow truly were disturbing the model - though your video doesn’t show it - I wouldn’t start with a deflector. Just turn off the fan in the filament profile. The deflector seems unnecessary when there’s already a built-in firmware solution. If you still suspect airflow, try the draft shield in Orca Slicer (possibly also in Bambu Studio). I dropped Studio back in January when BL adopted their customer-hostile stance against 3rd party support. Orca is better anyway. With Orca, you can generate G-code and transfer via Studio or SD card if your printer is locked down by Bambu’s firmware.
Keep in mind, draft shields are really a holdover from open-frame bedslingers running ABS or other temp-sensitive filaments. Still, it’s an experiment worth trying but one word of caution, if you don’t like filament waist in purge, you’re going to really hate the wasted filament using a draft shield.
From your video, as you pointed out I see wobble caused by nozzle drag across the model. Have you tried a different layer height than the creator’s default 0.08? The lower the height, the more “squish” the nozzle applies, which increases scraping and wobble. Raising the layer height reduces that effect - you should see improvement in the first layers to be sure. Honestly, unless the creator was going for some kind of “smoothness” affect, I don’t see much need for 0.08 here. It’s a lampshade. Chasing ultra-smoothness just weakens the structure and likely contributes to the fragility we’re seeing.
I ran a flow test print and found the exact same max flow as in the Bambu Studio PLA Matte preset (22mm³/s). According to the slicer, the flow is below that limit for this print.
I skipped the draft shield because it uses twice the amount of filament. Since the P1S has a fully enclosed printing chamber, I just turned the auxiliary fan off. I also increased the layer height to 0.12mm. The print quality was satisfactory until the head stopped for about 20 seconds. By staying still on top of the model, the hotend created a small ball of melted PLA that ruined the rest of the print.
I made this print with a brand new SD card. I have another one on the way that I’ll try later with the timelapse feature turned off, but I’m not very optimistic 
. Hopefully, I’ll be proven wrong!
Just be aware that if you use an SD Card with a capacity greater than 32GB, only the first 32GB will be recognized. In fact, the printer may not recognize the card at first unless you format it in the printer.
