The "Flying Extruder" Mod

27:1 is plenty gearing as it is. It has plenty of strength for pushing 3mm filament. There is no point gearing it even higher as the nema 11 only ever gets warm and doesn't overheat at all, all while having enough torque as it is. Gearing beyond 27:1 would just slow down your ability to do retracts.

I'll look into it sounds good.

Quote
DjDemonD
Sounds good I see there is a 100:1 nema 11 being offered also why not use that one? Would need much less motor current.

The motor speed and therefore the steps/mm would be very high. If your controller has variable microstepping or jumpers to set the microstepping, then you could reduce extruder microstepping to bring the steps/mm down to a value that the controller can handle. However, you might need more drive voltage than the controller can provide to maintain the speed. Using 24V power would help substantially, but even that might not be enough.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
o_lampe
The idea of the suspension is to enable the extruder to follow the effector moves without adding the mass/inertia to the equation.

How would you build a fixed platform between three carriers that move up and down independently?
-Olaf

Ah - I knew I was being stupid, but couldn't figure out how...

Thanks.

Quote
dc42

Quote
DjDemonD
Sounds good I see there is a 100:1 nema 11 being offered also why not use that one? Would need much less motor current.

The motor speed and therefore the steps/mm would be very high. If your controller has variable microstepping or jumpers to set the microstepping, then you could reduce extruder microstepping to bring the steps/mm down to a value that the controller can handle. However, you might need more drive voltage than the controller can provide to maintain the speed. Using 24V power would help substantially, but even that might not be enough.

Realistically, really, its going to retract _slowly_. Not good. Enough reason to avoid using it. 100:1 will just slow it down a whole heap when it doesn't need to be slower. up to you :-)

I which case I wonder how the flex3drive handles retraction as it is highly geared.

Edited 1 time(s). Last edit at 12/09/2015 06:15AM by DjDemonD.

I found one supplier offering this motor [www.ebay.co.uk] which is 350g with gearbox so probably not worth the effort to change over (and thats without the metal support bracket I'd need to fit it).
I also found this one [www.ebay.co.uk] (a smaller motor so less torque???) would it still be sufficient?

The one I'm using is this one: [www.omc-stepperonline.com]

I followed this design. Links for parts are in the description: [www.thingiverse.com]

The stepper motors themselves on the two pages seem to have hugely different holding torques. The ebay stepper motor claims 6Ncm ( 0.06Nm) min torque, while the link I posted claims 1.3Nm which makes me think their unit conversion is off by a factor of ten. Not sure what to make of it.

Thats the one I was looking at but the supplier says its 350g, have you weighed it and found it was actually 200g?

Yes, just over 200grams for the complete flying extruder, that includes the filament drive spur, screws and PLA. (209grams if I remember right). I'm in the middle of a large print atm but I can confirm the weight of just the stepper motor later if you'd like.

EDIT: Could be that the stated weight is the weight for postage?

Edited 1 time(s). Last edit at 12/10/2015 05:49AM by TTN.

No don't worry it sounds fine I just wondered where the discrepancy came from. Im getting pretty good results with my nema 17 flying extruder and have optimised the design for the printed parts so Ill try these before switching to this lighter motor at some point, thanks for the advice.

Quote
DjDemonD
I which case I wonder how the flex3drive handles retraction as it is highly geared.

Flex3drive seler responded to my questions:

• Maximum filament extrusion error due to the torsional deflection of the shaft is less than 0.011 mm.
• 0.7695 mm of filament is extruded per one revolution of the stepper motor. You want to retract in at most 0.1 s and you will need to retract about 1 mm. That means you need a stepper which can achieve 13 revolutions per second. That is doable; probably even at 12V (definitely at 24V because I'm in that range and at higher torques).

Posted my flying extruder on thingiverse [www.thingiverse.com]

Thank you all for posting your ideas and experiences with this system.

I finally put mine together tonight, here's a picture of it:



Haven't had much time with it yet, but I have high hopes. I've hit the limits of a long bowden tube, this one is 5 times shorter than the old system so that's got to be 5 times better, right?

Looking forward to seeing how it performs.

It will perform much better, perhaps 5 times is optimistic but what results were you getting before? Since going over to this setup my kossel is now my fine detail printer whereas before it was being beaten by my i3!

Results previously were pretty good, but not spectacular. Good enough for most purposes I guess.

But now...

Oh. My. God.



The one on the left is with the old bowden cable. The red one on the right is with the new flying extruder.

No cleanup on the pieces at all, these are straight off the build plate.

Notice any differences?

Edited 1 time(s). Last edit at 01/08/2016 10:21AM by nebbian.

Excellent. Might post a Marvin myself. It really does beat a long bowden by some margin. I've got a flex3drive to try which might do even better. Good filament is essential but with poor filament control it can't be used to get these sorts of results.

print something bigger and try to find resonance artifacts.

Hey Sungod,

I wasn't getting resonance artefacts before, so if they appear then I know where they're from. I'll certainly look out for them. I did notice the extruder motor wobbling around a bit, perhaps it needs some damping.

Can you post a picture of the artefacts you're seeing?

Its hard to say where this is coming from and its possible that its something else. Im working and a 2nd extruder that is complete bowden to test that. In addition the layer stacking could also be influenced. or at least im not happy with the stachking on my machine atm.

But there is an argument to be made the this coming from the swinging motions that the rubberbands are allowing.



Edited 2 time(s). Last edit at 01/08/2016 08:14PM by sungod3k.

Sungod,

That doesn't look like resonance to me, if it were then the period would stay the same as the diameter changes. From the picture, the artefacts follow the diameter of the piece.

This suggests to me that it's a model or slicing issue, ie not enough facets.

I hope this helps

Here is a marvin printed on my kossel at 0.1mm layer height.

Quote
DjDemonD
Here is a marvin printed on my kossel at 0.1mm layer height.

That's a nice Marvin, well done! I find white models very difficult to photograph well, the detail always ends up being washed out.

Are you using ABS or PLA? Do you have a layer fan? Your keyring seems better than mine, but I think the eyes on mine are better than yours. Hard to tell from the picture.

Quote
nebbian
Thank you all for posting your ideas and experiences with this system.

I finally put mine together tonight, here's a picture of it:

[attachment 69215 IMG_5147.jpeg]

Haven't had much time with it yet, but I have high hopes. I've hit the limits of a long bowden tube, this one is 5 times shorter than the old system so that's got to be 5 times better, right?

Looking forward to seeing how it performs.

any chance you release your flying extruder on thingiverse? Looks fantastic.

Quote
nebbian

Quote
DjDemonD
Here is a marvin printed on my kossel at 0.1mm layer height.

That's a nice Marvin, well done! I find white models very difficult to photograph well, the detail always ends up being washed out.

Are you using ABS or PLA? Do you have a layer fan? Your keyring seems better than mine, but I think the eyes on mine are better than yours. Hard to tell from the picture.

Yes it was a pain to photograph. This is PLA and I am using a 0.3mm nozzle, I have a mini fan of my own devising Kossel Mini hotend fan with duct, I would probably get even better results with a really powerful layer fan but I didn't want to add too much mass to my effector (probably wouldn't matter for this Marvin as I printed him at 40% feedrate - he took an hour to print). I think his eyes are less than perfect the overhang curled upwards a little during printing and was only partially flattened by the subsequent layers.

This (and the previous 3 Marvins) are my first attempt at high resolution (well medium I suppose) printing as I usually print larger objects requiring far less perfect surface finish. Also my seems were aligned with the left side at the corner of the eye, I sometime use random seem position but this tends to leave marks on the sides of the part, its a shame in slic3r you can't choose aligned seems and then specify where the seem should be on the model. Perhaps you can achieve this by rotating the model?

Quote
Rossini

any chance you release your flying extruder on thingiverse? Looks fantastic.

Thanks Rossini,

Here you go:

[www.thingiverse.com]


You'll have to bodge up your own brackets to mount the rubber bands on your carriages, I just used some thin wire I had lying around.

Quote
nebbian

Quote
Rossini

any chance you release your flying extruder on thingiverse? Looks fantastic.

Thanks Rossini,

Here you go:

[www.thingiverse.com]


You'll have to bodge up your own brackets to mount the rubber bands on your carriages, I just used some thin wire I had lying around.

Hi nebbian,

I want to say thanks for sharing your flying extruder design here. I tried it with my delta printer (150mm diameter and 170mm build height) last night.
It looks awesome, but sadly the effector joints came off when I move the effector to the edge of the build radius. Well, either magnetic joints are not strong enough to keep the bowden tube bent, or the delta rod are not long enough.

Here is the photo:

By the way, what's your printer dimension?

Thanks Rollingdice,

it's fantastic to see my design on someone else's printer!

I can see the problem you've got -- you've cleverly mounted the hotend a lot higher than most people do, which minimises effector tilt issues and gives you more build volume. But it also increases the leverage on the effector by the bowden cable.

That plus the magnetic joints are the issue.


I would try making your bowden cable longer, so the rubber bands are angled the other way, and are only there to stop the extruder from flopping around too much. This might mean that the bowden cable is more free to bend. I can see that you already put your extruder on top of the A frame instead of underneath it like mine.

Another option would be to run a spring from between the rods at the effector to each carriage -- pulling the joints together. Some people do this to reduce backlash in normal ball joints.


btw my build volume is around 190mm diameter x 170mm (I think). I've never tried to print anything higher than about 100 mm. I've come close to the outer limits of mine, however.

Edited 1 time(s). Last edit at 02/03/2016 08:23AM by nebbian.

Oh, so that's why people building delta printer with normal joint use spring on the rods. However I doubt this will help on magnetic joints.

Thanks for the response neb.

Never mind the extruder can I get a pic of the other side of the hot end? how are the cooling ribs integrated in the hot zone?