Is your Extruder Feed TUBE causing print defects?
Вставка
- Опубліковано 28 чер 2024
- Those tight bends MUST be causing print defects right?
Help support this channel! / needitmakeit
This one was suggested by a viewer and it's a question I've had for several years, does the friction inside the tube cause the extruder to produce an inconsistent result? Before starting the testing, I was pretty sure that with such a tight bend in the tube and with the tube in the cable chain, the extruder must be having trouble. But if that were the case, why would Creality design it this way, is it a bad decision or do they know something that we don't?
I hope you found this video helpful and interesting!
Another way to support this channel indirectly is through my affiliate links (where I receive a small commission). It helps me to continue to provide high-quality content videos. Thank you for your support!
NEW MIC (I bought this)
amzn.to/3vcMKAr - Rode Wireless ME mic (Saved me about 4+ hours/vid)
amzn.to/3NBFvZe - Mag clip (Super handy, but a little pricy)
BAMBU PRINTERS
shrsl.com/4aghz
CREALITY PRINTERS
shrsl.com/4aghy
CREALITY HYPER PLA
shrsl.com/4aghn
#creality
#crealityk1
#klipper - Навчання та стиль
Cool testing! We gotta solve all these mysteries to understand as many variables as possible.
Thanks 802, I'll end up doing some more extreme testing, I have a few ideas that came to mind after making the vid. How long of a tube is too long? 10' ? 20'? I still don't think 10 or 20' would cause problem unless it looked like a spring-back type air line.
Nice video - I loved that you measured how strong is the extruder pulling filament. BUT, maybe people could be having problems if they use filament dryer boxes behind their printers, so, long tube with friction of their own too. So, everything compounds to having problems.
Certainly it's possible that with the long tube and a double roller as well as a tight bend from the filament going into the feed tube sure. Here's what I'd like to do, I want to buy 20' of tubing and then run the test on multiple different printers, we can do with the stock tube and with the extremely long tube. Having that much tube should make it more pronounced and I don't know of anyone using a tube that long.
I got so sick of dealing with bowden tubes so i converted both printers that used this to direct drive and replaced any feeder tubes with 2.5mm ID tubing. So far it's worked exactly as I had hoped and I personally find it to be a superior setup. Luckily I had a lot of it already on hand as one printer is dedicated to printing with filament made from PET bottles and the diameter isn't exactly consistent... generally around 2mm. The 2.5mm ID tubing solved any feed issues I had so I applied this logic to all my other printers and I don't regret it. Any printers that had long feed tubes had them shortened to about 2" out of the extruder. I did this with my anycubic kobra 2 when I moved the filament holder to the top of the frame. Worked like a charm and this thing prints TPU like a champ.
Well done! Thanks for the useful data! Now I know for sure it's my original extruder causing the artifacts. I've got a new Microswiss hotend for my K1 as well as the new extruder and that should work really well. Much appreciated!
Good stuff as usual!
Thanks for taking the time my friend!
This is what I had expected -- Creality makes a pretty strong extruder, and the K1 with the right nozzle can really firehose out a lot of filament very fast. A little bend in the tube isn't going to be significant.
It seems that way for sure. I don't think it's a bad idea to reduce the tube length and drag as much as possible, it can't hurt, but we're not in a range where I'd be worried about it. I'll have to do some more testing with a 10' tube and see if that changes anything. Can I even buy a 10' tube?
@@NeedItMakeIt Maybe try with much longer tubes. Especially for today's bulky multi-spool filament dryers that might be too big to place near a printer. Maybe I like to dry my filament in another room lol :)
@@JameTek I was thinking about trying to get 10 and 20' and see what that's like. I still don't think it would be much of a problem, but I could be wrong.
Cool video, thanks for the work you have done, I was surprised that PLA was as touch on extruders as it is, BTW I got my exercise in today. I exerted my 72 grams of pressure and you got a like!
I've done some testing on this in the past and it was with a Bowden setup, I had similar results on that tests, PLA always seem to resist more, those tubes were tighter of course. Thanks for the like, it all helps!
I wonder if the tube, being snaked around and restricted inside the drag chain, is causing a pushing force on the print head, leading to some artifacts. You can see the tube has a couple of tight bends as it moves to the edge.
I had a cable bundle too tight on a direct drive one time and it caused the print head to move slightly from the tension caused by the cable bundle. I adjusted the cable bundle and the problem went away. It took me a long time to track down what was happening.
This would actually make more sense to me than restriction in the Bowden tube. Especially since not every printhead will be affixed perfectly or just as tight. We have seen this happen with the cables and tubes leading to the head on other printers.
I print a lot of ASA, and I noticed a large difference with PLA+. After a long ASA print (100C bed, 260C nozzle, cover in place, door closed), when I retract the filament for a change or storage, the ASA will have very obvious marks from the extruder drive hubs on the last inch and a half or so of the filament. PLA+, even after a long print with the cover off, barely shows any drive hub marks. That tells me temperature has a bigger role than you might initially think.
you do some real good tests keep the good work, i always thought how people stupid are when they tell you not to use capricorn or say that bend is restricting the feed of filament, always knew that force of extruder can over come this, only down side can be worn ptfe tube it will slice it from the inside of the tube over time....
Thanks, they're pretty low cost tests at least for now, but as I get more traction on YT, I'll spend a bit more on proper testing equipment, not that it's really needed most of the time. I'll be building a test rig for some of my upcoming video as well and I have a nice load cell so I can avoid adding weights, it should speed up the video a bit as well. I'm curious about the wear and why it causes more friction over time. Is that because the path becomes narrowed?
@@NeedItMakeIt over time i have seen on fb groups people posting pictures with hole made by the filament friction, same as it does on max rubing against the glass cover 🫢🫢only there filament does it from the inside altho it will probably happen if you print with alot of movement inside of the tube, different models different way ur extruder will operate
I agree with others when they say the Creality extruder is strong enough to handle that with no issues. So the PTFE tube can be in the chain or out. The only reason I took mine out was to reduce wear as you stated and also to help when printing brittle filament like carbon fiber-filled filament. I saw lots of breaks in the PTFE tube when the tube was inside the chain. And once I removed it I saw much less breakage. So in my opinion why not take it out and use zip ties. Great video
That's a good point about the brittle filament, I have never had this issue personally but if you don't have a dryer or keep your filament in the open air, it's going to begin to breakdown. I like that PLA breaks down over time, but I can't print the parts that I need to last with a filament that becomes brittle. I would assume that It also becomes more brittle in the form of a 3D print over time?
Great video!
Thanks, I'll probably follow up with a more extreme test when I have some more time. How about sending some filament through 10 or 20 feet?
@@NeedItMakeIt Oh the length of the filament is an interesting concept. I run 6 Bambu Lab printers and one thing I've encountered is having the filament wear through the PTFE tube over time. I found one P1P that I've owned for about 10 months that rubbed all the way through so you could actually touch the filament. Obviously that impacted print quality. I quickly decided that I need to replace my PTFE tube on a regular basis.
@@spoolheads Does anyone know why when the filament wears into the tube, the friction increases? is it that the filament path is always consistent and we then have a wear path that matches the filament diameter more closely causing more contact areas with that filament? I wonder how I can setup a test for this, and to see how long it takes to wear through using different filaments. CF I would probably wear the most, or maybe GF.
This morning I woke up to a print on a P1S that was totally under extruded. When I took the tubing off I could see that the filament had really worn through through most of the tubing at the entrance. There was a little U shape just the right side of the filament now. Seems logical that it could cause more friction.
Testing it seems pretty challenging. This PTFE tubing has been on the printer since I got it and the printer has something like 1500 hours on it. I’ve just been printing PLA plus. I’m sure more abrasive filaments would wear faster. I think I’ll start changing the tubing every 500 print hours or so.
So I don’t know if the tubing was causing the issue or if it was just a clog or both. But now when I have an issue with my P1S printers, I chance the tubing, change the nozzle and clean out the extruder. Just finished the first print after that it it looks a great.
Sorry for the novel here!
Thank you for doing this test. I can sleep better. ☺️
I was expecting to have some different results to be honest, but that's what the test showed, unless you're running Bowden and pushing filament to the hot end, it's going to be very hard to relate any extrusion problems to the feed tube (Some people call it reverse Bowden) I'd prefer to not call it that.
You are rising on my personal scale for 3D channels.
I would draw you a graph ... but, you know, YT :)
Nice! Little by little I'm getting there, I'm pushing hard for 2 videos/week, hopefully I can get to a point to be doing 3 or more. This tube issue is something that kept coming up in conversations and I think we can pretty well put it to rest. I have another idea for a test video which is really about taking the extruder to the limits, but I have so many videos planned now, I'll need to get better to be able to handle the videos at the rate that the ideas are coming in.
This video is awesome, but honestly... You should have combined those graphs to see how the values change in a single screen.
Shoot, I had them that way at first, but then it was really messy, good point though I'll take that into account for any future tests. Making those graphs is a bit painful if I'm honest, less graphs is better!
It's true you have 5 materials, but if the vertical bars are not adjacent but grouped per material, the extra space between the groups should make the graph nice looking.
Check the article by Prusa about annealing to see an example of multiple groups which still look ok (and you would have fewer bars in total than them)
The AMS unit in the Bambu Lab printers is pretty weak, and definitely has issues with filament path friction.
That setup loads and unloads the filament. The tight bend before the print head is the most critical point for loading failures. Unloading problems seem to be more difficult to sort out. Filament without a smooth finish adds to the friction and so is more problematic.
Possibly people are confusing those issues with simpler reverse bowden setups like the Creality K1 you are testing and so thinking the reverse bowden is a cause of problems. The AMS issues don’t affect print quality - they halt the print with a failure message, then you have to correct the problem.
I have the AMS, and I again haven't had any of these problems, not even a single time have I had a problem with filament failing to load. When switching between the side mounted spool and AMS, I've forgotten to re-connect the tube for the AMS, and that will lead to a lot of filament being fed in, but it never reaches the printer... Could this be related to the filament, stiff and brittle PLA filament? I know a lot of people keep theirs in the open. There are several connections between tubes as well, and I'd imagine that inside there is is a possibility to have filament getting hung up.
Why do you think that a rough surface texture case more friction? This seems to be true, however it has less contact areas and should cause less friction... the same concept as a BIG WIDE drag race wheel vs a small tire.
More investigation is required and I plan on doing a follow up video using 10' 20' long tube to see if I can cause the printer to produce defects because of the tube length.
You might look at whether the manufacturer actually used PTFE tubing or plain nylon tubing. My QIDI Tech X-Max 3 came with a blue supposedly Capricorn tube for the feed tube. I replace it with PTFE tubing that was the white translucent color and it feeds with a lot less resistance now
There is one way to know that's easy, if you cut it, it has almost a gritty feeling and also it makes a strange sound, it doesn't seem at all like cutting plastic.
Thank you, NeedItMakeIt!
Looking at your work I made a decision and K1C just arrived - do you know any good source for Prusa Slicer profile for K1 / K1C ?
I would switch to Orca slicer. It has built ins support and you can print direct through wifi.
Nice! K1C looks like a nice unit. I'm still happy with my K1, but there are some nice changes on the K1C that I wouldn't mind having.
I haven't looked recently for one, but when I did, I didn't find much out there. Like Alan, I went to Orca and haven't looked back, it's a very nice program and there are profiles for the K1 printers, I'd have to look to see if the K1C has one since it's fairly new.
Even if the print quality doesn't change it's always a good idea in a mechanical system to have as little drag as possible. The less drag you have the less work your machine needs to do. Which is always good.
Sure, that's a good rule. I see some people running with the long tubes coming from their MMUs, the tubes are in the several feet range, but they are fairly gradual in their curve at least.
I still want to experiment from a top feeding and more centralized setup, the tube would still be there, but would come in from above and be fairly straight, similar to a bed slinger.
I am finding that I am getting a lot of breakage of pla in the middle of the tube. This is with old pla but I suspect it is from the tube bending too sharply as described in the video.
Your old PLA is probably just wet. Wet PLA become exceedingly brittle.
Try drying it, and see if that helps.
Yes, that's definitely a property of PLA, it's meant to breakdown over time and it will become very brittle unless it's stored in an air-tight and humidity controlled container. As @ZappyOh says, try drying it out if you have a filament dryer, it may still be useable afterward. The chemical makeup of PLA changes over time.
I have a hemera based printer that has had problems with this depending on which filament and colour of filament. Different pigments have different lubricities. Black (carbon black is slippery so it has to be neutral or help) being more slippery and white (titanium oxide is an abrasive so it has to be neutral or add friction) being least slippery. I have had some filaments that are actually difficult to pull through the reverse bowden setup for no reason.
I'm sure that the surface properties contribute to some extent, from what I've seen it's more to do with how stiff the filament is as it tries to resist conforming and pushing against the walls of the tube like a spring.
It might be worthwhile trying to test a 10' long tube to see if it causes any problems, I still don't see it, but maybe at that length the friction will be high enough to begin to cause problems. From what I saw, retractions are a non-issue these tubes are large enough that I can push back about 4-6mm before I start to feel resistance. Pushing seems to be harder than pulling.
I've always wondered why the tube isn't textured on the inside to reduce friction. The texture would quickly wear off, or it would be too hard to produce?
I haven't finished watching the video yet, but wouldn't it also make sense to do a test where the filament spool is mounted above the printer so that the tube doesn't make that tight 90° bend just before entering the extruder?
Nice test but will the filament swell up when its absorbed a lot of moisture
From what I've measured, it doesn't seem to swell at least not swollen enough to be able to measure with a set of calipers, PLA filament will become brittle over time as it is highly affected by exposure to air (moisture) over time. This is something that people have complained about having their filament break inside the feed tube as it become too brittle.
Very interesting Video, didn' t thought this.
Btw. where did you got the two sloth models at 10:42 ? Like to have the STLs or a Link to them.
Just curious, was this a new piece of ptfe tubing or was it used? I just replaced my reverse bowden ptfe tubing and was quite shocked at the difference in how slick the new tubing was verses the well used one.
This tube was new, sure that's a good point, after some time of using them there will be wear and I suppose that the path of travel of the filament will be sized closer to the filament itself, maybe causing a bit more friction. I don't know of any rules of when to change the tube, maybe when they wear through? How many hours would be a good idea. It will of course be based on the type of filament you print with.
If that's the quality of the K1 Max, I'll stick to Bambu. Cool video though :)
Yeah, the K1 and Max aren't quite as good as the Bambu's as far as overall quality, I'm not sure if it's the extruder or the entire structure. They are at different price-points though.
Cool video, the k1 extruder is rated at 50nm so in theory should be able to pull just over 5kg max before the gears skip.
Your test showed less force than 50nm before skipping.
Good to see it only takes a tiny amount of weight to actually pull the filament through a bent tube.
I got my p1s yesterday in replacement for the a1.
First print out the box no settings changed..veroni rabbit...almost perfect quality.
I am reprinting the file standard settings with the same roll of filament on my k1 and my ender 3 v3 se to compare the difference.
Will be uploading a video later.
They are pretty shocking
Keep up the experiments your vids are cool
Could you make a video on printing with a 0.6 or 0.8 nozzle?
Please!
What is the opinion about CHT Nozzle sir
CHT seems to be proven, it's not just a gimmick. I have plans on trying the Bondtech CHT Volcano style nozzles against the HF ones that Creality offers but it'll be at least a few weeks until I can get to that. From everything I've seen and read, it's the real deal and makes sense to be able to heat the filament to the core more quickly. Are you aware of any that are available which are not made by Bondech for the K1 series printers?
@@NeedItMakeIt Thanks for your reply
You should have tried not attaching the tube anywhere. From the toolhead to wherever it terminates behind the machine. It would go up, eliminating 1 tight 90* bend, and no bends as it moves around
I wanted to revisit this, so maybe I can set something up in a future video.
I like this test, however I question how you have done the test. Filament is not pulled out of the tubing at a 90 deg angle (your test bed exit point), it is the reason why TPU is showing best results, it's able to flex more freely and therefore creating less far less friction at your filament exit point. If you were to do the same test with your test bed exit point mounted vertical you would see a very different result for all filaments.
Sure, the tests rig could always be changed. If you look to the test, the tube bends and that eases the angle, but I'm happy to provide a follow up test result. I thought it was better to test the filament as it comes through the tube to the printer in a horizontal fashion then coming down to the extruder at an angle.
Even with the absolute worst case, we're nowhere close to having extrusion problems caused by the tube. I think the only way to have more certainty is to take it to even more extremes and buy 10' of tubing and see what happens.
Of course if you're running a Bowden setup it'll have completely different results, pushing through the tube is never as good as pulling and tube diameter being tighter causes more friction.
@@NeedItMakeIt Thank you, I really like you videos. I’ve a K1max extracting filament from a dry box so my PTFE tubing is longer then a stock printer and I’ve not yet noticed any disadvantage. However, over time I know the tubing will need replacing because the lubrication properties of the tubing will eventually diminish, for this reason I stock a couple of meters of the tube.
@@Jamesn-js1zp It might be a good test to thrown in at some point, tube friction over time. It's a strange one because the tube itself is a homogonous material, I suppose the slippery-ness comes in part by the slick stock surface being intact, or maybe there is some type of treatment on the surface. It's a bit odd that it would differ so much. If you take a piece of nylon and remove the surface, it's the same below, why would the tube be different? Do you know the max length of that tubing I'd like to setup another quick test to see how much more friction there would be in say a 5' 10' 20' section. Seems people are interested in printing with long tubes.
@@NeedItMakeIt I wasn't a fan of how the filament was mounted on the rear of the K1Max, so using a filament dry box mounted beside the printer enabled easy access. My PTFE tubing is 1060mm long, plus there would be some friction pulling the filament from the spool sitting on rollers inside the dry box. Regarding PTFE tubing lubrication properties, as I understand it the PTFE lubricant is imbedded in the nylon when it's extruded, they mix PTFE powder with the extrusion lubricant to form a paste and it's used to extrude the tube.
1:25 wait, do you mean ptfe tubes from bowden setups are thinner by nature? (
Some people call the DD setup with a tube going to the filament a REVERSE Bowden, I'd prefer not to use that term to prevent confusion.
On a Bowden style setup, we have an extruder that feeds to the hotend with a tube between, on that setup, the walls of the tube are thicker and the filament has only a small amount of space to move around inside the tube, this is needed to keep the extrusion/retraction in an acceptable range. If we didn't have it, there would be too much for lack of a word 'backlash' within the tube to be able to accurately measure and compensate for.
The Feed tubes on a DD setup are thinner walled and have a larger ID, more space in the tube for the filament, but in this case it is irrelevant because the extrusion/retract happens at the hotend what happens before it is not so critical as it was in the other scenario.
Capricorn tube for example is the thicker walled stuff for a Bowden setup, it's not to be used for a DD setup. We're looking for something closer to 4mm OD and 2.5mm ID.
I hope that makes sense.
Issue isn't print quality. Issue is filament keeps breaking in the tube, including the PLA that shipped with the printer. Also kept dry and stored in a room that is constantly air-conditioned on dehumidifier setting
I can only speak to my personal experience that I have never once had filament break in the tube. I've heard of people having this problem. If that's the case, the seal was broken and humidity got in, give it a good long dry and see if that helps.
After 8 years of printing, using the least amount of feed tube produces better print quality, regardless of the printer. More consistent prints for longer and less maintenance and problems.
I'd agree when there is friction, it's always better to minimize it, In the case of a newer DD printer with a solid dual gear setup, you'd have to be into the 6+ ft range for it to start becoming a problem. Retractions are so minimal, we're mainly dealing with extrusion on these printers. If we're talking about Bowden, that's a completely different story.
I have an idea for a follow up test to push the printers to their limits, and perhaps it's a good test for any new printers as well. I'll be onto it in a few weeks, I think it's relatively straight forward... except things don't always go as planned
I'm sorry , but who is sourcing different ptfe for DD vs bowden, its all the same ptfe from the mfg. Except geeetech thunder with an enlarged outer diameter and the stifness it creates.
Bowden setup must be far tighter to the filament, 4mm OD, 2mm ID. Capricorn tubing is my reference for Bowden style tubing if you're curious. The DD setup has a 2.5mm ID, this is also confirmed by my Bambu printer which has the 2.5mm ID tubing and of course a DD setup. What we're sourcing vs what we're supposed to purchase as a replacement, that I cannot say, my Bambu came with replacement tubing, I have not yet had to use it. Creality did not send extra, I only have the stock tubing from the Max and the K1 to verify against.
The answers here show why Creality has issues and the users don't know how to fix it. Any restriction, no matter the power of the extruder, cause issues. Things are moving so fast resistance will cause problems.
Sure, it you try and take off like a drag-racer, you'll peel the filament apart, but that's a different problem. Bambu runs the same tube as Creality, the same length and diameters, it also bends quite tight if you watch. I have to stick by and say that at least from the observations and testing I've done, extrusion issues on Creality come from a different source than the tube.
If you have some data to share please do, I can always come up with a different testing method to see if we can prove/disprove.
Try a vertical tube route with +15% extra length from the furthest print head distance. Horizontal routing is the wrong way for how the printer is setup.
Depends on the printer, k1 max is all horizontal till the hot end, so they set up works
But why your print quality is so poor? :)
There are a few reasons, one is that I have lighting setup to make my prints look as bad as possible. In other words, I'm trying to show the defects. The spray paint also helps to do this. The parts are also quite small and I'm using a macro lens to pickup any blemishes. From the printer standpoint, I did not use Aligned seams, but random, and they always look bad unless you have your filament dialed in. I also used the stock setting for the PLA filament and I did not calibrate for PLA nor for flow for really good looking prints, I'd need to calibrate it.