CRITICAL: Missing Z-hop movements during printing are causing print failures and nozzle dragging. #3423
Comments
|
The 3D printer and the nozzle are straining, which could lead to damage to the machine. Please, fix this as soon as possible. |
amzaldua
changed the title
@amzaldua I saw your post and have observed the same thing. It is quite fortunate that you have "extensive knowledge of G codes and the kinematics of complex 5-axis machines" as you stated in the forum. The overwhelming majority of users would not have been able to debug as effectively at the gcode level. Have you already added an issue for OrcaSlicer as well? Thanks. |
|
No, I haven't done it yet, but since you suggest it, I will do it to help OrcaSlicer users. Thank you. |
|
Awesome. Thank you! |
amzaldua
changed the title
|
Wanted to add that I've experienced this problem as well. Relatively flat prints with a lot of infill experience this the worst, and the issue seems to be exacerbated with smaller layer heights. 0.24mm seems to avoid the problem, probably due to more generous tolerances, but 0.08mm always results in the nozzle scraping over infill, regardless of the type of infill. |
|
@amzaldua Hello, personally ,this place doesn't require a Z-hop. Z-hop only occurs during retraction and layer change. And after printing, I haven't encountered a print failure. Could you please tell me why a z-hop is needed here?
|
@XunZhangBambu is it possible for users to enable a Z-hop before crossing infill? While I agree that it's not needed under ideal conditions, it's possible to drag across infill if the infill is overextruded or if the print suffers from slight warping upward. The condition of the nozzle dragging across infill is so commonly observed that there is a nearly universal admonition on numerous Bambu-related forums to avoid grid infill whenever possible. This is troublesome for two reasons: (1) grid infill is the default infill in Studio and (2) Prusa users (and users of other printers) typically don't experience the issue. While I appreciate that skipping the Z-hop does speed up printing, it has resulted in a widespread recommendation to use gyroid infill to obtain quieter prints and avoid potential hotend damage. As you know, gyroid infill is one of the slowest infill styles available, but many users use it to avoid the grinding/scraping noise that occurs so often when using grid infill with no Z-hop over the infill. The minimal speed gain obtained by eliminating Z-hop for increased speed across infill is a disservice to Bambu Lab's many users. That slight speed gain is strongly offset by the many, many users complaining about "loud" infill, reports of scraping noises and users thinking that their printers are being damaged by "dragging across infill" when they use traditional infill styles such as the default grid. I can't imagine that this is the desired user experience for Bambu Lab printer users, but it's what I've observed as a regular user complaint. As I often point out, every user's experience is different and is strongly dependent on the user's skill level and environment. I'm sure you have many, many hours of experience with 3D printers, but that's not typically the case for most Bambu Lab printer users (or users of many other brands, for that matter). |
Disable this setting. |
|
I completely agree with the points raised by @tsmith35. I would like to further emphasize that another issue arising from this is the frequent occurrence of failed prints due to the piece detaching from the print bed for the same reason. Furthermore, the continuous friction between the nozzle and the piece raises concerns about potential damage to both the nozzle and the printer's mechanical components. I recommend that, by default, the option to simplify z-hop should be deactivated, and users should only enable it at their own discretion and risk. |
|
@SaltWei, by disabling "reduce infill retractions," the dragging is reduced, but only in the infill lines. The nozzle still scratches the print on the walls or in the internal solid infill. |
|
I had the same issue. I´ve got some improvement on my prints using rectilinear infill, and using z hop on my custom filament settings, and the extruder settings. But im getting the scratch nozzle issue frecuently, im not able to print more than one piece at same time because of that. Single piece often fails too, and i cannot use the AMS lite, because the purge movement its critical for this issue. Im sure that its firmware/software problem. People who has not this issue, often prints simple things like boxes without infill, little things, and not organic models. The machine firmware is the 01.01.01.00 video_2024-01-18_12-46-37.mp4 |
|
Ewww... I hate that dragging clicky sound. It sounds so nasty. What is that box of cast-away parts? |
More than 2 kilos of wasted filament because of that issue. I had 95% failures. I opened a ticket with bambu lab support, but i´ve got not answer yet. I really think that this machine is an awesome one, but the slicer/firmware is not finished, or we have defective machines |
One big thing that I discovered about 3D printing is this: no two users have the same exact situation. What works great for one person may not work at all for another person. If I had a dollar for every time a BBL customer complained about scraping / dragging / noisy infill / nozzle knocked parts loose / don't use grid infill ... well, I'd be able to buy at least a new X1C Combo with the funds. Everyone else in the world uses grid infill. Not us. It causes problems. 🙄 |
I used all type of infills available on bambu slicer. 2 kilos of tests can speak for me. The infill with less scrapping situation i found its rectilinear. I tried lot of configurations and all the calibrations available. The point is that the default profile, should be ones who works on all prints, even if the speed or the %infill excede the necesary, and after test it, you will improve settings to make it beter for you. I found that the A1 profile its radically the opposite. Its a hardcore profile for speed and very simple shapes, that will fail with organic forms or multiple pieces around the bed, or just tree supports. Lot of bambu users are just like the apple fanboys. Repeat the same obvious things thinking thad the people isn´t smart. It´s maybe projection on the others... |
This is caused by multi-reasons, most is not caused by z hop, but bad bed adhesive. You can try to increase the z-hop and see whether it helps. From my side, cleaning the PEI bed by water with detergent, and enabling brim does help! |
Do you mean that the z hop is still not enough? Please show your case and 3mf, we can have a check whether the z hop is generated. |
This is not slicer issue, this is caused by bad bed bad bed adhesive. Maybe one side of the wipe tower is wrapped. |
|
Nozzle dragging happening on my P1S too, especially when ironing. |
I am not developer of printer. I can only help to check whether z hop is normal or not in your 3mf file. |
|
I have given up on this getting fixed. Yes it is Software based, i came here from OrcaSlicer because to them this is a BambuStudio issue, there is a lengthy thread over at Orcaslicers github about the issue. To me however none of the mitigations worked a 100%, i can pinpoint that $something changed regarding slicing / code generation or whatever between OrcaSlicer 1.8 and 1.9 because this is when the whole knocking over parts issue started. @ZackySteel i have some 10h+ prints with large tall objects that will get knocked over in recent Orcaslicer versions. |
|
I used to have nice prints but more than 50% of my prints have failed that worked previously. This is multiple settings, multiple P1S printers, multiple filaments. There is great adhesion but if you hit a part enough times it will come off. Looks like I will be downgrading software versions. Hope that solves it. |
I'm at the same point except with the A1 and I only had it about two weeks before I had to admit to myself that something was definitely wrong with how many print failures and massive cosmetic flaws it kept running into. I am so baffled as to how there doesn't even seem to be any real communication from Bambu Lab on this as it affects every single print negatively and even if it doesn't cause complete print-failure, it's rare for my A1 to complete a print without very noticeable surface blemishes that in, in many cases, have been bad enough that I don't even feel comfortable providing that print to the client and instead try printing it again or call it quits and move to one of my older printers. It's so bad, I literally purchased a new "main" printer last night so that I'm not just up a creek when I need to get things printed/completed quickly. I had hoped moving back to 1.8.2 might be the fix for me, but unfortunately, despite copying all the settings that were applicable to a version that far back, I was seeing print-times increased by over 50% compared to the same models sliced on the newest versions. After wrestling unsuccessfully with that for a while only to realize upon attempting to print that I also couldn't use the AMS Lite (at least without further modifications that I wasn't aware of), I had to just call it quits on that front. I'd love to know if anyone has any helpful info on getting everything working well on 1.8.2 or on the actual base problem itself, because it's just a nightmare right now. |
|
This program changes the default infill type values and disables the “Avoid Infill Retraction” option for all profiles. There are several problems related to Z-hop and dragging: I’ve also noticed that in internal wall overhangs, Bambu Studio doesn’t detect them properly and fails to adjust the fan speed accordingly, which results in curling and the nozzle dragging on the internal overhang contours. Orca Slicer solves this issue with the option I’ll show you below. |
|
Oh! One last tip: To tighten the back screws of the hotend heating assembly, it’s better to first remove the layer fan. This way, you can access the four screws without fully loosening this component. It’s crucial to avoid disassembling the hotend heating assembly because the cables are very rigid, and it’s common for the joints to break. |
|
The older version of Orca slicer worked for me. I didn't try the different settings in the new versions because I just wanted the issue solved and it sounds like these are work arounds to minimize the issue, not the solution. Is that correct? |
|
I've noticed an increasing number of discussions and complaints from users in the past few months, all related to Z-hop issues. I just read half a dozen discussions (here and elsewhere) related to this bug. Any chance that a solution can be fast-tracked to fix what was broken? It's taken quite some time just to convince Bambu Lab that there really is an issue, thus there is a large amount of documentation showing that the problem does exist and that it causes both quality and productivity issues. |
Great point. I wish I would have known this a few weeks ago as I broke the thermistor wires attempting to tighten the screws and installing a new heating assembly was harder for me than replacing my heat bed with all the tight places and my fat fingers. |
Thanks for the tips! I only realized upon trying to open it that I guess it's a Bambu Studio profile and not OrcaSlicer? I see "Avoid infill retraction" and am still looking for z-hop (which I know I had found at one point but never decided to change), but was wondering if there were any other important adjustments that you might not have mentioned specifically. I'd love to know in order to see if I could get something mildly workable in OrcaSlicer! As for the internal bridging, the crazy thing is that the pieces I have been printing (using that term loosely, unfortunately!) are literally just wide hexagonal pieces, extruded straight upward for the middle part of a sword and yet, despite even cutting each individual part in half, hoping the height was the cause of the issue, I'm still having them consistently fail on the last few layers. One thing I did notice about my most recent attempt was that I thought I might have been hearing a very light scratching sound almost right from the start and noticed that unlike the previous attempts, this print had significant scarring on the first layer as well as the last layer. I'm just at such a loss, having come into the third or fourth of my most recent project's larger prints having been knocked off their initial place on the bed and seeing what should be the last few layers instead on the ground in a giant ball of string. |
|
Also, if anyone has created, worked with, or knows about an A1 profile for Cura, I'm pretty desperate at this point. My only other currently available option is literally a Makerbot Replicator 2 that refuses to print anything without massive vertical artifacts along every available vertical surface, so I'm looking for any potential option to actually be able to use the printer I paid hundreds of dollars for just over a month ago! (slight frustrated undertones) |
|
@ZackySteel try using an older version from April, hopefully before the z-hop bug was introduced |
|
Z-Hop settings can be found in filament settings -advanced when ur looking for them in Orcaslicer. Try Orca 1.8.2, for me thats the last version before these z-hop issues. |
I did try jumping back, but was pretty overwhelmed by first not even thinking about how of course I couldn't just carry my profile from the newer version over, but second, by how much longer each print was saying it would take, with my first model being over 50% longer. I had tried to adjust all the settings that I could, of course having to exclude those that only exist in newer versions, but wasn't sure if I was missing something or if it was just that much slower. That on top of how rough it is already losing so many settings I'm used to in Cura caused me to just put that attempt on the backburner and hope there was some way to get it working on the newer OrcaSlicer. Do you (or anyone else) have any 1.8.2 OrcaSlicer profiles that work nicely for you and have you noticed if it really is just a lot slower than newer versions, regardless? Lastly, on the OrcaSlicer front, I was actually just going to ask how you handled having two different versions at the same time, as I've only ever selected the "uninstall the older version" option when installing and was wondering about how to keep two versions without running into issues. I really appreciate your taking the time to give these tips! Thanks! |
|
I completely forgot one of the most important things I wasn't sure about! When I initially tried a print with one of the older versions that I had adjusted settings on, I noticed that when I was starting it on the printer, it didn't seem like there were options for using AMS-lite. I had seen initially that the older version I was using didn't have a set profile for the A1, so after realizing I might not be able to use the AMS-Lite, I was also left wondering if it would do its resonance-compensation and flow-checks either. Apologies for the long list of messages; I hope this is the last roadblock in the way before I'm able to at least somewhat use the printer again! I'll also see if I can't open a ticked with Bambu Lab, themselves, just to ensure that they hear about it and hopefully realize this is a real and serious problem. Thanks again for all the help! |
|
@tsmith35 @ZackySteel : A related PR is fixing the missing zhop. After internal discussion, we will merge this PR #4631 |
|
I believe I am facing this same issue. The printer head has now started scraping the print and also leaving extrusions on surfaces in particular around sharp corners. Does this sound similar to everyone else’s experience? For context the printer worked flawlessly for the first week and now this week everything it prints has this issue no matter the filament type or how much I calibrate settings. The fact that this seems to happen for people a few weeks after having the machine makes me think it’s an hardware issue of some sort that only kicks in after so many hours of use. Adding images as a reference. |
Hey there! Thanks for the heads-up! Does that mean I'll need to use another page for further comments and/or questions? No idea how that works on each issue's threads when the two are merged. Also, I have seen that this is an issue not just related to missing Z-hops, but missing retraction movements as well. Has that also been noted and just something that will also be fixed as a side-effect once Z-hop is fixed? I run into issues with this even when I don't use Z-hop and I know I've seen someone's analysis in which they find that the G-Code generated is missing certain retraction movements. Any update on whether retraction movement is also being handled in the same fix as Z-hops and what kind of timeline might be realistic for a fix would be greatly appreciated, as my A1 is unfortunately still acting as a paperweight, at the moment! Thanks so much! |
Hey there! I know I've seen information from someone posting on this problem about actual missing retraction moves (which I kind of assume might also cause missing Z-hops, if said Z-hops are supposed to occur on retraction moves) and that definitely made sense with what I'm seeing. I think the most difficult problem with issues like these, especially when we're trying to diagnose them right after purchasing a new printer, is that we often spend the first few days messing around with test files (which may or may not be sliced with the same problematic release of whatever we run into the issue using) and then on top of that, I know for me, it took a while before I felt I could definitively say: "Yes, this is a real problem and not a fluke." and "No, this isn't something I am causing or can fix with more setting changes, cleaning, nozzle swaps, etc." My A1 could technically print smaller things, but I would hear more and more scratching noises and see worse and worse scarring as I continued to print more as well as larger models. At one point, I noticed that there was filament sticking well out of the nozzle after a print and that upon starting a new print, it wasn't able to be wiped off by the usual process the printer goes through, which then caused it to make scratching sounds even against the first layer it put down. So, in short, I definitely think this is a tricky slicer-related problem that, for the reasons I mentioned above, presents itself in a very difficult to diagnose manner. Last, your issues definitely look similar to what I've been going through, especially with top-layers (the pain!). Hopefully we'll get word on a fix here in the next few days, because I only really got to have my "fun new printer" feeling for a few days! Good luck! |
|
Perhaps what it's happening is that after some time scratching because of the slicer bug the hotend block is displaced or moved a little and that aggravates the problem. |
|
This thread unfortunately discusses various different causes as one and even for Bambu Lab developers, it is hard to tackle them individually. I authored #4631 and can offer my opinion: Your problem seems to always be the nozzle scratching on surfaces and/or knocking over prints, specifically with taller objects. Cause 1: Reduce infill retraction. As already said, with filament cooling down, uneven surface structures will occur. This cannot be prevented physically. Reducing retraction when moving to infill also means not doing a Z hop, and this ultimately will cause the scratching on uneven surfaces. Try disabling this first. Cause 2: Travel distance threshold. This setting's default is 1mm and it means if the distance between the previous extrusion and and the beginning of a new extrusion is less or equal to 1mm, no retraction and no Z hop will happen. This also was mentioned, so try to set it to 0. However, please also be advised that to frequent retractions for short amounts of extrusions will lead to increased filament temperature and hence also can cause stringing. Note: the first two causes are what this GitHub issue is originally about, and Bambu Lab addressed this. The following causes might be your problem but are not related to this issue. Cause 3: Infill pattern. Some infills are known to be prone to scratching, especially the default one. Try using gyroid infill, it prevents lines crossing each other. This is also well documented and caused by filament cooling and shrinking or extending depending on the structure. Cause 5: Over-extrusion. Extruding too much will lead to uneven surfaces and can also lead to nozzle scratching or knocking over prints. Try to follow one of the well known methods, I recommend printing a hollow cube with two walls and then measuring the width of the wall to recalculate your flow. This is well documented. At least one of the pictures in this thread looks like over-extrusion. Cause 8: Z-hop height. Try to increase the Z hop height, this could help to not touch other print perimeters when they might have lifted up after cooling down. Be aware, this can increase stringing. Cause 3: Poor build plate adhesion. If the build plate adhesion is weak, prints can lift, and even if with all previous mentioned causes addressed, it will lead to the same problem. I noticed that some of you already checked this but I wanted to mention it nonetheless. Clean your built plate and try using glue. Cause 6: Bed levelling. Usually, Bambu Lab printers will detect this before printing but nonetheless, some of the posted pictures definitely look like quite unlevelled beds. Can be observed best with the first layer. Try to run a full bed-levelling via calibration settings. Cause 7: Z-hop type. Primarily related to stringing and filament residue at extrusion end. Happens when a normal Z-hop type is selected (vertical lift). Try to use spiral Z hop, it is the best lift mechanism when doing a Z-hop and this is also documented in Bambu Wiki. However, there is a bug with it on layer changes which the PR I created will fix. I hate writing walls of text in GitHub issues but I sincerely hope, this can help you. I know a lot of this was mentioned already. Please don't get frustrated, most of the times print problems are caused by one teeny-tiny detail. I finished all of my tests and calibrations and my A1 prints absolutely flawless (as long as I compile from the branch with the PR #4631. |
I wasn't sure if this was a reply to my question on merges or not, so I wanted to reply just in case! I've gone through pretty much every possibility I can think of, but (and I don't know if Bambu Lab workers are already aware of this), there was a guy who posted somewhere, showing that there were actually missing retraction movements in the generated G-Code. There are definitely various ways to try to work around this issue, but it's been a rough deal when I can't print as large as my Ender 3 V2 Neo and can't even come close to the immaculate surface finished (I really love smooth surfaces! XD) that my Ender 3 V2 Neo, Ender 3 Max, Two Makerbot Replicators, and even my little Tina2 that could all can produce. It's doubly frustrating when everything is so closed off, so as to make using other slicers like Cura as difficult and as non-optimal as possible, without being allowed to use vibration compensation among other optimizations if not printing a file sliced with Bambu/Orca. I understand wanting to have one's own software, but, having jumped into the slicer race so comparatively late, it's very frustrating to see all the efforts to close customers off even while they will of course be stuck dealing with the lack of features and optimizations that longer-standing competitors' software have (not to say that they aren't getting there, but instead that this is simply the reality when competitors have such a long head-start). I'm mainly just worried because I know that the retraction issues topic was already well over eight months old with no communication from Bambu until just a few days ago when it was merged into another issue that seems significantly less dire in-terms of the impact it has on those affected. Especially with such concrete documentation and examples of people literally going through the G-Code showing where retraction moves should have been generated but weren't, I'm very confused as to how we've received no real word on timelines for a fix or even communication confirming that Bambu Lab is aware of the problem. Apologies for the long message! Hopefully we'll hear from them soon! |
|
@ZackySteel @Waylango could you post your .3mf project files here? I would like to have a look. For @Waylango, ideally for the two pictures you posted and for @ZackySteel the sword you mentioned. Try to export the last setttings you've been using. Can't promise anything but I tracked down most of slicer bugs with gcode analysis. |
|
@ziehmon im in the process of setting up my printer closer to my pc to make trouble shooting a little easier. I feel like files printed straight from the Bambu handy app print fine but when I print them from Bambu slicer they have these issues. As soon as I’m set up I’ll run another test print and send you the .3mf file from one of the prints I shared. Thank for the help and info folks. Very very informative. Chat soon. |
Ah! Thanks for mentioning .3mf; I'd completely forgotten about that! The settings I (hopefully) exported it with should be nearly if not identical to those used on the print itself, the only difficulty being that I was using a laptop and the printer at work before and had to download Orcaslicer on my home computer as well in order to export it here. Real-time edit, here: Unfortunately, it doesn't seem .3mf files are supported either, but it let me attach it after sending it to a zip folder. Please let me know if you run into any issues opening it and I'll try to get on it as soon as I can! Thanks again! |
|
@Waylango Your issue is related to this ticket, but it is a different issue that can most likely be solved in a different manner. Firstly, the overextrusion in the corners is a good sign that your k-factor or dynamic flow calibration is not working properly. Make sure to perform a k-factor calibration or run your A1/A1 Mini once with the Dynamic Flow Calibration on once. Secondly, the other issue is "Infill/Wall overlap", which causes overextrusion specifically between the solid infill and the walls. I notice that your print only has overextrusion inside the walls, but likely between where the infill and walls are. This kind of overextrusion is caused by this setting. I have mine set to 0%, but if you do that, you should just be aware that your flow ratio needs to be tuned properly. This often happens with concentric infill, which it seems you are using. Finally, make sure your flow ratio is tuned properly. It should be below 1.0. If you are using arachne, I highly recommend setting it to never print lower than the nozzle width. That means make your minimum wall width 100% in the arachne settings. From what I can tell, your issue is probably not related to arachne, as you don't seem to have it enabled. It looks like you are having issues both with your k-factor/dynamic flow calibration and infill/wall overlap. I believe in bambu studio there is only one infill/wall overlap for both solid and sparse infill, so just make sure your sparse infill is still attaching to the walls correctly after making that change. Solid infill will print just fine with it set to 0%, while overall having less issues. |
|
@vadixidav thanks for the input. After playing around a lot in Bambu slicer I seem to be getting much better results. I’ve tried a million things so that makes it hard to pinpoint where I was going wrong. But things are looking a lot better now. I’ll keep all your feedback on hand when I run into issues again. Attaching some images of my latest project and as you can see it’s looking a lot cleaner even with this cheap two trees filament.
|
|
@Waylango for future reference, I would recommend visiting other community threads like the official forums, Discord, Reddit, etc for improving your print settings. GitHub issues are mostly just for developers to receive technical information about bugs and feature requests and tend to be focused only on the issue you are currently on. In this case the issue is specifically related to scraping during travel, which you likely experienced, but you can find help for your other issues elsewhere in the future. |
|
@vadixidav yip I’m on the official forum but I can’t post anything there for some reason and the discord has also blocked me from getting in because it says I’m using a vpn even though I’m not. Not the best experience I’ve had with a discord forum I must say. |
|
After the last PR we continue with the same problems. Any update? |
Bambu Studio Version
1.8.4.51
Where is the application from?
Bambu Lab Official website
OS version
macOS Sonoma 14.2.1
Additional system information
Apple M1 Macbook
Printer
Bambu Lab A1 series
How to reproduce
Printing a G-code generated with Bambu Studio with a non planar superior surface.
I attached an example.
Actual results
The nozzle scratch over piece.
Check the movement between 43684 and 43685, there is no a Z-hop movements.
Expected results
The nozzle should be retract in Z axis between movements in the same layer and in the same type of line. Always.
Project file & Debug log uploads
debug_Fri_Jan_19_08_40_03_43376.log.0.zip
scratching example.3mf.zip
Checklist of files to include
The text was updated successfully, but these errors were encountered: