Ultimaker2 Feeder System - Improvements and Ideas

[rick 0]

yeah i wish i could show you the insides but i'm printing a crossflow fanoutlet atm, so it's probably not advisable to open it up right now.

Also i was thinking to replace the bowden with the same spring ( i have 5M of it) and see what it does. If the spring can stay compressed under the pressure of the filament it will probably be a better choice than the plastic tube

 

[rick 0]

Rick,

I'm pleased Ian can see it, but the rest of us lower casts don't have permission to view it. Have you made it Public?

 

I have now ! sry about that, didnt know i had to put it on public.

 

[rick 0]

Oh and to let you guys in on what i'm working on... Here's the idea.

I'm working on a system that feeds the filament trough a threaded tube(metal), For this to happen the directional feed of the motor has to change from vertical to horizontal. This is all good until i realised that the feed of M3 thread id 0.5mm per rev and the bolt feeds +/- 25mm.

So i need to either make the motor 50 times faster or find some sort of gear setup that doesnt take up to much space. I have tried worm gears, but the immense torque needed for the gear to drive the worm is not something i want for this system.(they're usually the other way around)

BUT ! here comes the fun part. If this works, you will have absolutely zero slipping because it will be physically impossible to slip through a cm or so of thread. Also this will solve all wear and tear problems with the feeder, as there is no contact with the actual feeder. And it will also be no problem feeding soft filament, because the threads will grip this just as easy as any other filament.

Just a quick rendering and wireframe:

 

So please share your thoughts and ideas on this, all help is welcome

 

Rick

 

[IRobertI 520]

My initial gut reaction is, can the filament resist the twisting force? I feel like there will be a lot of friction putting a rotational torque on the filament so that it will want to spin?

 

[rick 0]

My initial gut reaction is, can the filament resist the twisting force? I feel like there will be a lot of friction putting a rotational torque on the filament so that it will want to spin?

 

This crossed my mind as well, but tbh i don't think it will. M3 thread has a module of .0.4mm or so, this means that if u use ~2.85mm filament it will only cut into +/- 0.25mm of filament.

 

[takei-naodar 0]

This crossed my mind as well, but tbh i don't think it will. M3 thread has a module of .0.4mm or so, this means that if u use ~2.85mm filament it will only cut into +/- 0.25mm of filament.

 

Only problem i see though is, there is the same problem as with the first design of the feeder. What would happen if the PLA got a bit thicker lets say a bit over 3mm like 3.05mm after some months of storing it at home? Or even just 2.95mm? That would definitely lead to problems as there is no space for the material ,of where the thread has cut in, can go....

Another Problem as far as i can see in your last picture is you will need to use a gear onto it and there will arise quite some design problems on how to get the threaded part to stay in place and in touch with the driving gear.

With the Thread itself i personally would go for a special one ..... otherwise this idea would theoretically bring many problems in everyday use.

I personally would go for a M3.3 or 3.4 with p=0.7.

 

[IRobertI 520]

Well, I mounted my first printed prototype. It didn't help with under extrusion, but I didn't expect it to. Upping the current to 1500mA solves that issue for me so at this point I'm designing for convenience and because... why not.

It also confirmed for me how awkward it is to have the closing arm on the left side. Oh how I wish those side panels didn't stick out like that... My previous design went around that issue by making the housing fatter and the arm as thin as I could possibly make it (making it weak). Mounting on the right lets me make the arm fatter and the body slimmer. Trade-offs...

Another disadvantage to mounting it like this is that when you're feeding in material the natural curve of the material will cause it to press against the knurled bolt making it ever so slightly more annoying to load. When mounted the other way around the opposite is true, the curve keeps it away from the bolt.

There's a couple of other issues as well that I need to work around. Well it's one issue in two places, namely the mounting points for the arm and the latch. You can't over tighten those screws as it will constrain movement, but if you don't they will eventually wiggle loose. Meh.

I can also say that my UM2 is officially broken in now as I put a nice big scratch on the back of it hehe.

Right, time to stop rambling on.

 

[IRobertI 520]

Well, it doesn't quite live up to the renders but it's not too too far off I don't think

I'm getting close to being happy enough with it to release.

 

[skint 61]

Looks like a top job Robert.. good work!

I am wondering... is anybody actually 100% sure a new feeder system is in need of? A lot of people now seem to be printing highly successfully with the standard issue feeder?

This is not a criticism of all the excellent work in this thread by the way!! far from it.., its a genuine question of who actually is hoping and waiting for a killer new feeder mechaism ?

 

[pikey 0]

Looks like a top job Robert.. good work!

I am wondering... is anybody actually 100% sure a new feeder system is in need of? A lot of people now seem to be printing highly successfully with the standard issue feeder?

This is not a criticism of all the excellent work in this thread by the way!! far from it.., its a genuine question of who actually is hoping and waiting for a killer new feeder mechaism ?

 

I believe the current stance on this isn't a matter of IF, but a matter of when you'll have feeder/hotend complications with the current feeder design.

 

[chrisr 18]

So are the designs of Ian, iRobertl and others, flawless working replacements of the original UM2 feeder? Or do they also skip?

 

[IRobertI 520]

My latest feeder did not fix the issue for me no. (increasing the current to 1500mA did though. Unsure if my feeder helped as well) Like I said, at this point I'm designing it because I don't like the current feeder much. With my feeder the filament doesn't touch anything but bowden (unless you put on the extra guide) so there's no risk of ground down ABS making it to the nozzle and it makes changing filaments more convenient IMHO.

I really should get off my ass and release it for testing...

 

[cor3ys 28]

I really should get off my ass and release it for testing...

 

Yes i am waiting to get my hands on your design, I am currently running Bas from ultimaker test feeder which stops ABS particles going up the bowden, But i like the look of your quick release system and the extra attachment which might let me use flex material..

Have you tried it with ninja flex like material ?

 

[IRobertI 520]

I haven't tried ninja flex yet. I have a small sample I was gifted and I'm waiting for him to get back to me on some nice default settings to rule out my own incompetence with the material (never tried it before).

 

[chrisr 18]

IRobertI, on 08 Apr 2014 - 6:37 PM, said:

My latest feeder did not fix the issue for me no. (increasing the current to 1500mA did though. Unsure if my feeder helped as well) Like I said, at this point I'm designing it because I don't like the current feeder much. With my feeder the filament doesn't touch anything but bowden (unless you put on the extra guide) so there's no risk of ground down ABS making it to the nozzle and it makes changing filaments more convenient IMHO.

I really should get off my ass and release it for testing...

Ok, so does your feeder with increased stepper drive current work perfectly with no skipping?

I'm starting to wonder whether a feeder mounted directly on the head is a better design... I know that would slow up printing by adding weight to the head.... but I guess those types of printers with that type of feed are much less susceptible to skipping..... not sure if that could be done on the UM2 though. Hmmm.... :-|

 

[ian 32]

There are other ways of driving filament with zero skipping... just sticking with the existing ultimaker2 thinking wont do that.

I have a very good zero skip system designed.. tested and printing and its very different.. actually totally different to what ultimaker do. I was not happy with the inital concept of the feeder system so i went back to stage one and worked from there.

I am going to release my design in the big Fabcom 3D Fair (Germany) in my Presentation next month.

But I am also working on a smaller more conpact lighter version to be fitted to the top of the hotend.. then we dont need any bowden anymore..

It will be a very very interesting presentation I think ;-)

did anyone not notice I havent posted any photos in the new prints tread recently... LOLL... all my secret work every night...

Ian :-)

 

[Nicolinux 288]

Man I whish I could go to the Fabcon 3D Fair. But that's too far away from my place :/

They don't have a live stream by chance? Would love to see your presentation.

 

[chrisr 18]

Ian, on 08 Apr 2014 - 7:36 PM, said:Ian, on 08 Apr 2014 - 7:36 PM, said:

There are other ways of driving filament with zero skipping... just sticking with the existing ultimaker2 thinking wont do that.

I have a very good zero skip system designed.. tested and printing and its very different.. actually totally different to what ultimaker do. I was not happy with the inital concept of the feeder system so i went back to stage one and worked from there.

I am going to release my design in the big Fabcom 3D Fair (Germany) in my Presentation next month.

But I am also working on a smaller more conpact lighter version to be fitted to the top of the hotend.. then we dont need any bowden anymore..

It will be a very very interesting presentation I think :wink:

did anyone not notice I havent posted any photos in the new prints tread recently... LOLL... all my secret work every night...

Ian :smile:

Getting rid of the Bowden tube altogether is I'm pretty convinced, the way to solve this issue. So next question.... Is your new design, built, tested and proven that it completely solves the skipping problem?

(I'm just trying to figure out whether there is a proven and working design which 100% solves this skipping problem at this point in time as I will be having to make some decisions fairly shortly :-)

 

[ian 32]

well i will keep things secret until next month but i am printing evey night and day at the moment with my design.. im doing a 30 hour print at the moment with zero skipping.....

I dont think there is anything 100 percent with 3d printing.. but my design is... damn good ! ;-)

Ian :-)

 

[chrisr 18]

Ian, on 08 Apr 2014 - 8:06 PM, said:

well i will keep things secret until next month but i am printing evey night and day at the moment with my design.. im doing a 30 hour print at the moment with zero skipping.....

I dont think there is anything 100 percent with 3d printing.. but my design is... damn good ! :wink:

Ian :smile:

I'll take that as a maybe then :-)

Last question - are there any other of the alternative feeder designs in this thread which 100% solve the skipping problem? Apologies for drilling down for a simple "yes or "no" answer but I'd like to know - as I'll be having to make some printer decisions very soon!

Thanks. Chris

 

[ian 32]

Chris.. im sending you a PM now

 

[IRobertI 520]

Right, I've put up my design now. There's a long ass bit of text to read there so read it before you start printing. (so you can decide if it's worth the effort).

https://www.youmagine.com/designs/alternative-um2-feeder-version-two

Now we wait for Ian's design to make mine obsolete and pointless

 

[ian 32]

robert.. there are different ways to skin a tomato ! ;-)

 

[gr5 2,172]

The UM Original doesn't skip so obviously one can make a design that doesn't skip.

I haven't modified my UM2 yet and it prints just fine. But the ABS feeder *does* get ground up and bits make it up the bowden which is annoying so I plan to switch to a new design some day. No rush.

 

[IRobertI 520]

Curious to hear if anyone else has given my feeder a try? 3Poro reported that printing with flexible materials showed a big improvement for him.

Is it the wall of instructions that scare people perhaps?