Torgeir

The top hat was the simple way to extract filament fume out from my experimental room. So I had this plan, but before I could do this myself, my son fixed all this cause his mom did not like the fume from the PLA.. There is a fan that can be speed controlled and the big housing go to outside vent. So easy.. Torgeir.

Thank you, again going to the young ones.. They want the fume outside.. :)

Hi Folks, This topics almost went into the drain, -but actually this is a very important issue for an "FDM Bowden type" printer. When I had to tag this topics, I did not find the "tag" I was looking for, -so I just put up the "one" I have. But the fact is that, this is valid for every Bowden tube printers that's made. So you may ask, what is the issue; -the issue is friction! So, I'll say, friction is the Achilles heel of "FDM Bowden type" 3D printers. Friction in the transfer line some place; -"between the the bowden tube and the filament"! How to fix this? 1) Can we reduce the friction(?): Yes or No. 2) Can we increase the feeder torque(?): Yes or No. If you'd ask me in the beginning of my 3D printing "period", I would say no to the first question and yes to the second question. But today, I'll say yes to both questions! I'we put much effort into this problem, cause there was never any good answer to this issue, but now I understand the strange feeding we sometimes being observed during first layer and elsewhere during printing. Actually, I'd feels like walking around the bushes in this issue, but not anymore.. If this interest you, read on. I'll focus a little about the general things first. I’ll like to focus a little on the bowden tube and how it should be set up on this kind of printers. This might look strange for some, but may help others.. The bowden tube work best when it can stay with looping upward during printing, without tilting much to one or the other side. In order to keep this loop as perfect as possible, I’ve installed a tiny carbon fiber rod (max diameter 1 mm). This carbon rod is to be strapped to the wire bundle and only one attachment to the bowden tube at the very top of the loop. Like this: How the straps is arranged. Here you see the one place they are "attached" together. This is needed as both (Bowden and Wiring) are attached offset from each other at the extruder and at the feeder, now they can move freely without interfering each other. On the printer: And with a top hat special.. This rod should be solid strapped to one of the four screws that’s holding the extruder attached together. When referring to parts in the printer I’ll always refer to looking strait into the printer, -so the long screw to use for the carbon rod and the wiring wiring loom (or cable), will be the aft. R/H screw to strap into. This carbon rod also need to be strapped alongside the feeder unit together with the wiring loom. How to strap up the wire loom and the carbon rod from the feeder and toward the extruder is also an important detail to include in the whole setup. This strapping when done right, will prevent the wiring to be bend all the time in the same area when extruder move around, such thing is awaiting a rupture here (–an open wire) causing a malfunction in your printer to come. Also important to note is this; there should be NO connector located inside this wiring loom! Now to the question why do such a setup improve our printer? In order to explain this simply, just cut two peaces each about 1 meter from a roll of filament with a light color, white will be the best choose, if you have. Now you should have two equally curved up filament lengths, right? Find a black filth pen, the permanent type (no withe board pen, we don’t want remains from this into our bowden tube). No mark the outside of the loop in the whole length with a black visible line on one of this filament and make a black visible line inside the loop on the other piece of filament. Now, let it dry.. OK. As the “ink” is dried, we can make a little test. You do not need to power up the printer, but if you have an old UM2 you need to power up and just feed the filament to the extruder and stop here, do not turn the heater on (no need). For those with the +model, just feed, with the press wheel retracted, all the way to the extruder. Make sure you feed this piece of filament in the same attitude, as it is delivered from the roll of filament normally attached to the printer. Here is the two test filament, pre cutted deg. and ready to be inserted into the printers bowden tube. Make sure that the extruder is parked in the normal rest position, in the forward R/H position. Now you can choose any of the piece of the two test filaments you have, just feed it into the bowden tube as described above until it is at the extruder. Now look at where the black line is located when looking at the bowden tube, it should be possible to see this line. So, if you choose the one with the black line outside the loop, you will see the line following the upper side of the bowden tube. Right? And of course, vice versa, when feeding the other filament into the printer. So, as the filament extend into the bowden tube, the “arm” will extend while filament gradually change direction and this will increase the torque that’s will twist the filament starting at the feeder. So sure, this “twist” torque will variate as the extruder move around on the heath bed. So what is the point of mention all this, well, we can very much benefit from this “facts” -armed with this information. The key word here is friction. In order to keep the friction as lo as possible, we would like to have the filament as clean as possible when it is forced toward the upper part of the (Teflon) bowden tube. All this is to prevent additional “drag” for our feeder system in order to be as efficient as possible. If we can achieve this, we do not need this “very strong” feeders, that can create a force up (10-14.. and some more!!) KP., -a pressure “normal” retraction cant cope with.. Now lets have a quick look at the system: This is something from many months ago, however, an important note here is; If the friction constant is 1, we would not be able to move the filament inside the bowden tube at all. If it is zero there will be no resistance when moving the filament. Sure this is just the extreme extends (theory), a friction constant will always be some place between (0 and 1). According to this drawing, it is oblivious that outer “signature” track wheel cause increased wear inside the upper part of the Teflon bowden tube. Knurled wheel much less than hubbed wheel, so hubbed wheel should absulutely not be used on the outer side of the filament to drive the filament through the bowden tube! This because the hubbed wheel makes crossed cut along the whole filament length and really scrape into the Teflon. It is just a matter of time before it will fail, if it is installed grabbing the outer side of the curved filament. For the knurled wheel, this work better, but will in time also suffer for this kind of wear and tear. This occur when the system using knurled wheel have to produce higher pressure, the small pyramided dig up a nub behind every hole it makes and those small nubs work as small grinders to the Teflon surface. The Teflon surface is kind of soft and is easy to rub down. Printing this way is not very efficient as it create other problems due to the higher working pressure. The other feeding way, feeding in the inside loop, do not need high pressure, due to lo friction -and is much more easy to control.. Now to the worst problem here; when the bowden tube is “rubbed up”, due to this above issues, -you do remember the twisting of the filament during extruder normal movement. Well, -due to this twisting when the head move around, the friction factor might/will variate and so will the extruder pressure as well. I’ll guess you know very well how the rest of the story will end. Well folks, this one is not very easy to realize, but really is an improvement. Thanks. And thank you @gr5 for fixing up my mess.. :) Torgeir.

Of course, I'll say a 100 deg Celsius to be used for normal use, at last.. But what kind of "thermoplastic candidate" will this will be, I'll sure do not know.. However, we have a mixer, that's reduce the hot water to about +60 Deg Celsius. The focus is on the use of PLA in general for this issue. Thanks Torgeir

Congratulation @kmanstudios, Well done, as always - way to go.. Regards Torgeir

Hi Guys, An outstanding thread! Thank you. Regards Torgeir.

I could not agree more, -but also true the 400 step/rev will suffer for torque at a speed 1:3.11 (if my memory is right) .. :) More speed for a high number of coil stepper, means less torque.. with all respect. Torgeir

Hi undertherain, If you just upgrade from "the real UM2" to the UM2+, without all the changes needed, you might have a reveres feeder as a result (plus feeding much to fast). The thing is; an upgrade to UM2+, also ask for a new feeder installation. Yes, you have to change the whole feeder unit into a new one that is rotating the other direction. There is also a different speed for the feeder as this also will feed very much to fast(to much feed).. You need to change the feeder unit with the one needed for this software change, or you have to change the number of feeding steps and the direction of feeder back to revert back to standard UM2 settings. Thanks Regards Torgeir.

Hi jofleck, You can use the old print head together with an Olson block.. The feeder stepper servo used in the 2+ is a 200 step/rev type. Thanks Regards Torgeir

Hi Kmanstudios, Thanks for the flowers to the guys. :) And yes, PLA is the filament that's seems to be more and more used for medical issues. Just have a look here: https://www.ingentaconnect.com/contentone/ben/rpgm/2014/00000004/00000001/art00004?crawler=true And here: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5220021/ I'll think we're just in an early stage here. Thanks Regards Torgeir

Hi Folks, I've never been posting any prints (model) I'd made here, but this time I've simply have to share.. Right now I'm working in Japan, far away from home in Norway. So I have no access to my printer, but my two sons have.. Anyway, the older one sent me a picture of a print he made. To my surprise he made a 3D drawing of the old Hansa faucet we have on the kitchen sink, that's have a shower in combination with the faucet. This one is now kind of old, since 83.. so about 35 years I'll say. Anyway, the plastic housing of the shower part does not work very well any more and spreading water all over the place, -kind of awaiting to be replaced.. So yesterday I've got a picture of the old shower device together with a gray printed copy he made as an replacement for the old one. Not just that, he also sent a small video of how the new shower adapter worked. So he told me he is looking for some filament that's food qualified and can be used for producing this kind of parts for drinking water. He used PLA that's the stuff doctors using for support inside "our" body during surgery. The PLA dissolve inside body and leave no trace. -Is there something I'm missing here? Anyway, I'm impressed by those young guys, they just makes those things easily like that. Some says, seeing is believing. Well, another way of creating valuable parts using a 3D printer. Sorry folks, not allowed to upload a 1 Mg byte video, so had to skip. :/ I've almost skipped the whole thing due to this latter.. Thanks Regards Torgeir.

Hi Folks, Honestly, is'n this what can be done directly in Tinker-Mark2-extended-dual-17.09 firmware. So why using other methods to change any of the steppers direction? This is seen in one of the pictures in the setup menu for this firmware brought in here by klassie just a few mail back? This one: Does not this mean change motor direction? Or do I miss something here? Please check this one out in the advance setup menu. Thanks Torgeir.

Hi Gambit, Did you check that the "other directory", the one provided by @ghostkeeper, the directory under "roaming" containing another cura directory? This one: C:\Users\\<your username>\AppData\Roaming\cura\<Cura version>\cura.log (?) Thanks Regards Torgeir.

Hi Gambit, I'll think you need to uninstall the the 3.4.1 version of cura, make sure this "erased" directory is clean, then reinstall cura version 3.4.1 Thanks Regards Torgeir.

Hi Folks, Here is the path to cura config files in the c disk. C://users/"your name"/AppData/Local/ Roll down until you see cura. Copy all those files and delete those files in this directory. Thanks Regards Torgeir

Hi Folks, Excuse my ignorance here, never seen this eminent firmware before. But one thing I’ve notices is that this firmware has the possibility to inverse the direction of all axes including the two extruders. Why not try to inverse the X-axes in the menu? I’ll ask this question as both of you have the head moving the right side instead to the left where the switch is located. Just my two p. Thanks Regards Torgeir

Hi Folks, What a great analysis done here, lots of job with all this.. @gr5 I've been following this for sometime as this is kind of strange.. Just a note on your last print in here, the print you stopped before finish of first layer. At first I thought the beginning of «filling in» was OK. However, when copying this picture and zoomed into it for more details in here, I can see that the squished first line has a leak track, meaning the feed rate is to high. At the end of this, the nozzle do not leak any more -it looks flat as it should be. Maybe the bed is just a bit closer to the nozzle and really close the outflow. At this point the filament will sure be grind-ed, and this part of the filament will slowly move into the bowden tube. You now start to see the blobs, -that’s escape from the nozzle and you see how squeezed the last part is before a total lack of extrusion. Also those two blobs seems to be burnt, but may be a visual error.. When the prints go on, more and more of the grind-ed filament is feed into the bowden tube and this is become a serious problem for the rest of the print as the feeder have to work against the additional forces due to friction. So this is why your print suffer for under extrusion, as the grinding will continue.. IMO. The look of your nozzle also tell a story.. When everything is right, the nozzle looks great.. You should probably lower the first layer feed rate and also reduce the feeder 1 stepper current as this will prevent high torque when pressure is too high at the feeder side. This is one of Gunnar's suggestion in his long check list. Below you can see two areas magnified from your picture. Anyway, good luck. Thanks Torgeir.

Great mayo77, Keep up the good work. Thanks Regards Torgeir.

Hi Kmanstudios, This is what I tried to explain/said here for a looong, long time, but to "dead" ears, -I'll think.. I'm running the track line inside the loop and use less retraction due to less friction. I'd love to show some numbers about this, -but loose interest., ?... I can print things at quite a good quality due to those small modifications I've done to my printer. And it is still a plain "UM2" configuration printer, using the original setup -but have some small adjustment to behave proper.. What you show here should be corrected long time ago. There is a few of comments about this issue in this forum. Keeps up the good works. Thanks Regards Torgeir.

Hi There, You all are talking about the new Ultimaker S5 made in U.S.A. ? My telephone is a S5, maybe I could have another S5.. Hmm.. Take care. Thanks. Torgeir.

Sorry, but forgot to mention that you have to swap two wires on the feeder stepper motor, as it have to turn in the other direction. Either, pin 1 and two, or three and four on the stepper connector. But this is very easy to do.. The thing is; your printer will be (act) as a standard UM2 printer, so no firmware change! If you can improve your 3D printer with software, this is good, -but if you can improve your printer by hardware -it is even better! Thanks. Torgeir.

Well, -where I been working (in the aviation) we used 242 on those screws -3 mm (and the equal inch standards sizes) all the time.. I'd never had any problem removing those screws.. But, -if you fill all the coil with this stuff, there might be a problem. - - - So; we do not glue them with Locktite, we just lock them with a small droplet.. The black set screw is a high torque set screw, and the knurled wheel have just a few coils, -so here it is wise to use 242! Just my two and half pence.. Take care and good luck. Thanks Torgeir.

Hi fergazz, Thanks for the flowers.. Yes, I did many test and found that I do not need the Bondtech feeder. I'd actually made a test using double knurled and double hubbed feeder wheel. I found that the friction increased (of course) but the stepper had to be geared to a higher ratio as well in order to handle this extra "drag". It actually worked well, but I did not want to change the firmware from the original UM2! So I installed a 200 step/ 360 deg. and made a belt gear using same belt as the short UM2 for X/Y axes. The pull eye's I'm using for the gear is (16 and 32) tooth type 2GT. By doing this, no firmware change and your feeder now have close to double torque. This feeder unit make very little noise, just the humming from the stepper motor. Meduzas feeder unit is having the same gear ratio as mine, just go for it, but one very important thing to do is to mirror the original feeder housing. So to the question, why turn around the feeder? Ok. then, when we buy new filament from a shop, do you think we will have the one just arrived from the producer? If they have this filament in the shop, you'll have the oldest one, for sure. Meaning there will be some memorized curved filament for you to use, and even fresh filament have some effect -like this. Try this; mark the inner loop side of your filament (in same length as your bowden tube) with black (permanent) marker. Let it dry and manually feed this marked filament into the bowden tube. See that the filament line up its curved path into same curve as the bowden tube has, yes the black line will stay mostly on the inner (lo) side of the bowden tube. The black line represent where my knurled track line is located on the filament! On Ultimaker setup, the trackline is on the top and this make "some" added friction. Using hubbed drive wheel add much more friction in such setup.. I've always wonder why Ultimaker choose the outer side of filament to put the drive (knurled) wheel, but no one ever answer this question.. :( So I'll think Ultimaker always have a super fresh filament that's behave different, than "some aged filament" we the users got from shops around wherever we are. My printer can push approx 5.2 Kp input to bowben tube.. Then the bowden lock often just come off.. If you need this much force to drive the filament, there is some serious problem with your 3D printer. If you like to see my setup, I can share this -but need to be home in order to send pictures files etc.. Anyway, keep up your good work. Thanks. Torgeir.

Well, 242 is the blue one for screws that can be reopened. Some other Locktite product can really fasten to much.. Anyway, good luck. Thanks. Torgeir.