Is this an insane way to mount extruders?

I'm planning a core XY around a monstrous Diamond hot end. I've started to think about where to mount the 3 Bowden extruders. The print area is 300 x 300 (320 x 340 absolute max) and to get that I've ended up with frame which is 600 x 600. So, it was looking like I'd end with hellish long bowden tubes. Then I got to thinking that the shortest path to reach all 4 corners of the build area would be with the extruders mounted above the centre of the bed. I started looking at some sort of gantry arrangement when it struck that it needn't be fixed - especially as I'm planning to use V slot to build it. So, that's when I came up with the crazy idea which is denoted by the yellow bits shown in the pics. There is a gantry plate which can slide left to right on the front and back rails with a bit of V slot sticking up. Another length of V slot is connected between the two. The 3 extruders are crudely modelled, just a cuboid the size of a Nema17 with a lump on the side and cylinder hanging down representing the Bowden tube. These would be mounted on a plate with Delrin wheels and will run back to front. It's a passive system. That is to say that it is not driven in any way. The extruders get dragged around by the X carriage but only for largish print moves. For small print moves, they would be static - the movement being taken by the flex in the bowden tubes. It may need some sort of coupling between the x carriage and the extruder mount rather than relying on the bowden tubes which might get yanked out. As it is depicted, the tubes are about 150mm long which ain't bad for a 600mm frame. The whole thing could be dropped and the tubes made very much shorter still but then one would have to move the entire mass for even the smallest print move. There is probably an optimum height but I'm not clever enough to work it out.

I guess the nearest analogy to this is the flying extruder arrangement that Delta printers can have, but in this case the extruders don't need to move in the Z direction, I guess I could call it "floating extruders".

Anyway, let me know what you think. Is this just the deranged ramblings of a crazy old man? Got to go, there's a man in a white coat knocking at the door......

I'd say not insane, but complicated. I'd be worried about the bowden tubes being pulled out, but as you were thinking, some way of doing strain relief by attaching them to the carriages might be ok. If I had a Diamond, I'd design a custom 3 block extruder and just go direct (geared or direct drive) on them, and know that I won't be going super fast. The only tubes I'd have would be just as guides from the reels to the extruders.

Edited 2 time(s). Last edit at 06/07/2016 02:58PM by FA-MAS.

The whole point of using bowden extruders is to reduce the moving mass. Your extra gantry to hold the extruders just increases the moving mass beyond what simply mounting direct drive extruders would.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

It is an interesting idea but your build area is not that big.

I think an outer frame of 600 x 600 for 320 x 340 build sounds big but lets go with those numbers, lets also assume that the bed is in center of that. You now mount it on the length side with 340. Lets call that length Y

That gives you a maximum distance from extruder to hotend of:

X: 600/2 + 320/2 = 460 mm

Y: 340/2 = 170 mm

L = (170^2 + 460^2)^(.5) = 490.4 mm

So 490.4 mm will be the absolut longest straight length from extruder to hotend - Then give another 100 mm for bends and you come to ~600mm. I am not sure that is too much.


As for the concept of the idea: I think for a, say, 500x500 printer it would make sense to suspend the extruder over the center of the bed instead of on the side - But I would not have it moving. If you want to move the extruder you might as well just bite the apple and mount the extruder on the carriage it self. This is not Delta printers.

I think I remember someone talking about using two extruders for same filament line. One extruder to push the filament through the bowden tube and one very small extruder on the carriage, at the hotend, to push the filament the rest of the way.

I'm working on a single extruder coreXY design. I'm thinking of mounting the filament spool inside the printer on a swiveling bracket suspended from the top cover. When the extruder moves around, the spool will be free to swivel to follow it, minimizing drag on the filament. That makes the machine taller, but the filament spool has to go somewhere.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Yes, all good points.

Lets stick with Bowdens for the moment. I'm leaning towards 3off E3d Vulcan extruders. I guess it would be possible to mount these on to the Diamond but it makes what is already a pretty ungainly thing into a something huge, both in terms of physical size as well as moving mass.

So, attached pics show how it might be configured with the extruders static but in the centre of the bed, approx 200mm above the carriage. The X carriage is shown positioned in the extreme front left corner (ignore the simulated belts - these don't move in my OpenScad model when I change the XY coordinates for the hot end). The cylinders representing the bowden tubes are 300mm long so I reckon in real life they could be 350mm which is not too bad. I'm a bit concerned about what happens to the tubes when the hot end is directly under the extruders. They'll need buckle and bend a lot when the distance to the hot end reduces by 150mm to 200mm. That might induce some twisting forces to the carriage, although it's riding on 2 linear guides and should be well constrained. It's still a bit of a concern though. Now if only the extruders could move...............

It seems to me that you're basically looking to implement the flying extruder that's popular in delta printers.

I recently outfitted my delta with one: picture in case you don't know what I'm talking about

...and the spring mounted suspension might keep your issues at bay. The extruders are free to twist and change height based on head position, but don't have to be moved around laterally at all.

I've been thinking about trying something similar with my core xy variant.

Quote
deckingman
I'm a bit concerned about what happens to the tubes when the hot end is directly under the extruders. They'll need buckle and bend a lot when the distance to the hot end reduces by 150mm to 200mm. That might induce some twisting forces to the carriage, although it's riding on 2 linear guides and should be well constrained. It's still a bit of a concern though. Now if only the extruders could move...............

For this exact reason we don't do that setup on Delta printers. But delta printers also move upwards in Z height so towards the end this twist that you are afraid of become pretty severe. Still...

I will tell you a story from my first printer design;

Basically I had read somewhere that tightening the belt is really important - So I thought, how can I do that in a really great way? I came up with the idea of mounting a block that would tighten the belt like a ratchet wrench. It consisted of a number of parts working together in a very complicated way The time I spent designing it in CAD was maybe equivalent to what I spent on the entire Y carriage(!).

I never managed to print it. In the end I realized that tightening belts can be done by just pulling on the belt by hand and tightening it when the force is OK. Takes a little longer then with belt tightening screws and all, but it is afterall something you just do once...


MY POINT BEING - As I understand, this printer is also your first self-designed printer - Maybe just do it like the rest of the world? Stop putting time into this subject, put the extruder at the edge of the printer like everybody else. If you later find that the difference in bowden length ( 300mm to 600mm ) makes such a big difference, then sure, you retrofit it. But right now your goal is to get your printer off the ground. And honestly - Think very careful about the carriage design you are making. It is cool with the shower door wheels instead of normal linear rods/linear rails - But do they have to be installed like you do it now? Takes up a lot of space? That is the kind of questions you should be asking your self right now... Not if all the existing 300x300 mm printers with bowden got it wrong...

Edited 2 time(s). Last edit at 06/07/2016 08:34PM by LarsK.

LarsK, I hear you. However, I have a working printer with Diamond hot end grafted on to it so I'm not impatient to get this thing up and running. I'd much rather spend time getting things as close as I can at this design stage, which costs nothing apart from time, than start cutting expensive extrusion only to find that things won't work because the design was flawed to start with.

The shower door wheels are Open Build Delrin wheel kits, sliding on OpenBuild VSlot rails which is a lot lighter than round steel rods. My current printer has 2 rods, one above the other with the Diamond hot end hanging over one side. The whole thing flexes and twists alarmingly, hence the two rail design with the Diamond hot end slung in the middle. The Y carriages are Delrin wheels fitted to Open Build Gantry plates with a printed part bolted to it.

I can't do this like the rest of the world because AFAIK, no-one has grafted a triple extruder Diamond hot end on to a CoreXY design printer. If they have, I'd love to hear from them.

My goal is to have a greater build area, especially height wise, than my current Mendel variant and I want to make something that does not flex and wobble about but is robust and reliable. Secondary to that and because the build volume will be much greater, then the printing time will be longer. So, if I can increase the speed as well, it will be another advantage hence the reason for looking at coreXY designs. I may end up scrapping the whole idea and building a Delta instead.

Why is that whenever one asks a question about aspect (a), people always come back and say you should not be doing be aspect (b) the way you are. This whole thing is a work in progress so the entire carriage arrangement could end up being completely different. I'm just looking for thoughts on how best to mount the extruders.

Quote
maso
It seems to me that you're basically looking to implement the flying extruder that's popular in delta printers.

I recently outfitted my delta with one: picture in case you don't know what I'm talking about

...and the spring mounted suspension might keep your issues at bay. The extruders are free to twist and change height based on head position, but don't have to be moved around laterally at all.

I've been thinking about trying something similar with my core xy variant.

Yes, we are on the same page. The difference is that the extruders don't need to change height relative to the bed position (the bed will move in the Z direction) but the bed dimensions are bigger so XY movement will be greater. The spring arrangement sounds like it might be viable or a pulley/counter weight system which would keep the "tension" constant??

The thing is, I need to mount the extruders somewhere so if I can figure out a way that will keep the bowden tubes as short as possible, it'll be a bonus.

The Delta style flying extruder approach is the best for this and I can't really see a way around it. You will have to place the extruders in the center of the -supposedly- square bed and have the bowden length set to a value that would allow printing on the whole bed in one go -imagine a cube encompassing the whole bed- and for that you need a radius of the circumscribed circle to your bed's corners.

This basically means to have the vertical members of the frame high enough to allow a flying extruder.

If you make the extruders a sort of sliding extruders (I have imagined this with the extruder attached to a scara arm that swivers around one of the members) and have them shorter than a value that would allow a full circumscribed circle to your bed then if you print big you will get at some point the printhead to have to pull the whole weight of the extruders and that will affect the print. Even if you make a sort of counterweight system where the pull will be easier on the printhead, the difference from running free and then all of a sudden having to pull the extruders will be significant.

Edited 2 time(s). Last edit at 06/08/2016 04:41AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Quote
realthor
The Delta style flying extruder approach is the best for this and I can't really see a way around it. You will have to place the extruders in the center of the -supposedly- square bed and have the bowden length set to a value that would allow printing on the whole bed in one go -imagine a cube encompassing the whole bed- and for that you need a radius of the circumscribed circle to your bed's corners.

This basically means to have the vertical members of the frame high enough to allow a flying extruder.

If you make the extruders a sort of sliding extruders (I have imagined this with the extruder attached to a scara arm that swivers around one of the members) and have them shorter than a value that would allow a full circumscribed circle to your bed then if you print big you will get at some point the printhead to have to pull the whole weight of the extruders and that will affect the print. Even if you make a sort of counterweight system where the pull will be easier on the printhead, the difference from running free and then all of a sudden having to pull the extruders will be significant.

Yes, that's what the "Plan B pictures that I posted earlier show. Basically, with the extruders fixed (not sliding) about 200mm above the hot end. According to my rough and ready OpenScad mockup, bowden tubes of about 350mm at an angle of about 40 degrees will reach all 4 corners of the print area. For info, the bed as shown is (planned at the moment) 400mm square although the printable area will be less. I was just wondering about a type of design to alow the extruders to rise up when the hot end is directly below them, and so absorb the 150mm shorter bowden path? Basically no X and Y movement but some degree of (counterbalanced or sprung) Z movement just to absorb the bowden tubes when the hot end is at the centre of the bed.

The bowden loop has to maintain the largest radius possible to keep low friction between tube and filament. So best approach I see is having the extruder suspended from the bar, just like with a Delta flying extruder so it can swing in all directions, then somehow spring loading the extruder so that when the printhead approaches the position under it, it will rotate around its X axis bringing the bowden loop up... I will make a sketch later if this is too confusing of an explanation

Edited 1 time(s). Last edit at 06/08/2016 07:59AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Quote
realthor
The bowden loop has to maintain the largest radius possible to keep low friction between tube and filament. So best approach I see is having the extruder suspended from the bar, just like with a Delta flying extruder so it can swing in all directions, then somehow spring loading the extruder so that when the printhead approaches the position under it, it will rotate around its X axis bringing the bowden loop up... I will make a sketch later if this is too confusing of an explanation

Yes, I think I see where you are coming from but a sketch would be good if you have time. Remember that I have 3 bowden tubes, roughly 120 degress apart in a circular direction but I guess the principle your suggesting would still work? Thanks for taking the time.

Quote
deckingman
The thing is, I need to mount the extruders somewhere so if I can figure out a way that will keep the bowden tubes as short as possible, it'll be a bonus.

The shorter the tubes get, the stiffer they get. With 3 of them connecting the extruder gantry to the hot end carriage they'll be pulling the extruder gantry around as if it is all one piece. Unless you make those tubes very long (making them more flexible) and the printer very tall (to accomodate the long tubes), the moving mass will be higher than if you mount the extruders on the X axis carriage and use direct extrusion. That's why it's usually done with the tube(s) making a big loop at the top. Flexing a big loop keeps the load on the hot-end carriage pretty low, but then you have the extrusion problems associated with long bowden tubes to deal with.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Here's a 30seconds sketch for 1 extruder...



I couldn't wrap my head around the 3 extruders ... No time atm.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Quote
the_digital_dentist

Quote
deckingman
The thing is, I need to mount the extruders somewhere so if I can figure out a way that will keep the bowden tubes as short as possible, it'll be a bonus.

The shorter the tubes get, the stiffer they get. With 3 of them connecting the extruder gantry to the hot end carriage they'll be pulling the extruder gantry around as if it is all one piece. Unless you make those tubes very long (making them more flexible) and the printer very tall (to accomodate the long tubes), the moving mass will be higher than if you mount the extruders on the X axis carriage and use direct extrusion. That's why it's usually done with the tube(s) making a big loop at the top. Flexing a big loop keeps the load on the hot-end carriage pretty low, but then you have the extrusion problems associated with long bowden tubes to deal with.

Yes we've kind of got to that point. The latest plan is to have the extruders fixed (not moving) in the center of, but above the hot end. If they were (say) 200mm above the hot end it's looking like tubes of about 350mm will reach all corners of the build area. We're just trying to figure out if there is a way to use something like a Delta flying extruder type arrangement to take up the slack in the tubes when they are directly over the hot end, rather than having big or sharpish bends in the tubes. I guess I'm just looking for the best compromise arrangement which will give me the shortest bowden tubes without them exerting any great forces on the hot end or without the hot end having to drag the extruders around. Long tubes will cure one thing but cause other problems. Short tubes will do the opposite. What's the best compromise? I'll probably end up testing various ideas with bits of wooden rod, string and duct tape ..........

I have some thoughts on this based on my flying extruder setup on the delta that should work if your prepared to have some vertical height above the printer.

I have around 200mm of Bowden length with 580mm diameter coverage withe the extruder suspended centrally about 300mm below the central retaining point 500mm above the hot end. The central point is then pulley counterweighted to take up the vertical motion on the delta an this could be used to manage the small amount you need across the XY movement

This keeps the Bowden tubes short at the expensive of some central vertical height.

I'd mount the 3 extruders in a triangle arrangement with the motor drives facing the centre giving a balanced flying arrangement, I've even mentally designed the flying extruder mount

Quote
aussiephil
I have some thoughts on this based on my flying extruder setup on the delta that should work if your prepared to have some vertical height above the printer.

I have around 200mm of Bowden length with 580mm diameter coverage withe the extruder suspended centrally about 300mm below the central retaining point 500mm above the hot end. The central point is then pulley counterweighted to take up the vertical motion on the delta an this could be used to manage the small amount you need across the XY movement

This keeps the Bowden tubes short at the expensive of some central vertical height.

I'd mount the 3 extruders in a triangle arrangement with the motor drives facing the centre giving a balanced flying arrangement, I've even mentally designed the flying extruder mount

Now we're talking. Vertical height isn't an issue - the damned thing is (planned to be) a metre high anyway so it'll be floor standing. That gives me another 1.3 metres or so before I hit the ceiling. Actually, thinking about that, I guess there's no real reason why I couldn't suspend the extruders from the ceiling. My "man cave" is the one room where I don't have to get permission from "she who shall be obeyed". I think I'm pretty safe to assume that I can continue with the rest of the printer design and sort out the extruder mounting at a later date.

Here is a concept for a 3 way nema17 flying extruder mount that is suited to the E3D Titan...... first version allows for 48mm long motors which is excessive but I wanted to see the print size, 141mm x 131mm so will fit on most beds. For shorter motors it will be smaller.


I was going to angle the motor mounts in but remembered the Diamond hot end has the cold ends angled out.

and a version for 32mm long motors


Edited 1 time(s). Last edit at 06/09/2016 10:32PM by aussiephil.

Man, you are a star! Don't suppose you have that as an stl do you? No worries if you don't - I can cobble something together in OpenScad. Thanks for taking the time. I really think this, or something very like it, is going to work. Compared to my existing Mendel, my planned printer could end with 60% bigger X and Y, 500% bigger Z but with 30 to 50% shorter bowden tubes. Brill !!

Quote
deckingman
Man, you are a star! Don't suppose you have that as an stl do you? No worries if you don't - I can cobble something together in OpenScad. Thanks for taking the time. I really think this, or something very like it, is going to work. Compared to my existing Mendel, my planned printer could end with 60% bigger X and Y, 500% bigger Z but with 30 to 50% shorter bowden tubes. Brill !!

Just redesigning it so it's fully parametric, should have something in the next 24 hrs

It's been a learning experience in Fusion 360

Cheers

Ok think we are done......

1. Top view... this is the base for printing


2. Underneath with dummy 48mm deep nema17's to check layout


3 as an STL in S3D ready to be sliced


S3D tells me it's under 100gms at 50% infill

Here's the STL: (Removed for the moment, error in one hole size, working on it)

The three M4 holes for hanging should be positioned correctly for weight balance

I'll print one before throwing it up on thingiverse

Cheers

Edited 1 time(s). Last edit at 06/18/2016 01:44AM by aussiephil.

Man you are a star! I've never played with S3D or Fusion 360, only Open Scad (and I'm not very good at that) but I should be able to import your stl into OpenScad and play with it if I need to but it looks fine as it is. The other thing that struck me is that I need to get wires to the hot end and of course these could follow the same route as the bowden tubes which would be quite neat and tidy. So that's another problem solved before I even thought about it. Thanks again.

Quote
deckingman
Man you are a star! I've never played with S3D or Fusion 360, only Open Scad (and I'm not very good at that) but I should be able to import your stl into OpenScad and play with it if I need to but it looks fine as it is. The other thing that struck me is that I need to get wires to the hot end and of course these could follow the same route as the bowden tubes which would be quite neat and tidy. So that's another problem solved before I even thought about it. Thanks again.

The hole in the middle is to run your stepper wires along with any heater and fan wiring for the hot end. The only thing I'd like to do yet is to add some slots above the stepper motors.

I do plan to print one copy then I'll quickly do three variants for different motor lengths if it prints ok and is solid

Only started with Fusion 360 this week.... Been a step learning curve

Edited 1 time(s). Last edit at 06/17/2016 08:12AM by aussiephil.

Ok last edit for tonight

with some pretty slots in the top

Unfortunately I've broken the full parametric scaling, so no STL yet

Ok Printed it in PLA just to check out it would print and have found two things.
1. made the holes for the nema face plate to small..
2. could shave a mm off the top plate thickness, but not sure if the few grams actually matter.

oh and PLA sticks like the proverbial to a blanket on IPA wiped down Blue Tape, nearly destroyed the print getting it off the bed.

Just working my way forward on the design in F360 again after breaking the scaling last night.... except a slightly different look

The SD card is on top for size comparison

That's looking really neat. And small too, I imagined it was going to be bigger. I'll probably end up using short motors.

Ref getting prints off, I print PLA on borosilicate glass covered in Kapton tape (heated bed at 55 degC) just because that was how the original RRP Mendel came and following those build instructions. I've not got around to trying anything else and I've become quite proficient at applying the tape to the glass but it was a real pain to start with. I aways give the bed a wipe over with Acetone before printing. To get the print off I remove the glass and let it cool. Then, holding the glass on edge between my feet, I get a gasket scraper on one corner of the object and smack it a few times with a hammer! (Don't try this at home folks). Scares my misses to death for some reason - can't think why.

Quote
deckingman
That's looking really neat. And small too, I imagined it was going to be bigger. I'll probably end up using short motors.

It's hard to get a size idea from design screen shots or even photos without a reference

Ok, short and long motor versions now published via Thingiverse
[www.thingiverse.com]

Happy enough to publish this version.

I can feel a Diamond Hot End in my future......

Just thought I'd share these pics of what I ended up doing. Many thanks to AussiePhill for the idea but also apologies for not using his STL. I needed to tweak the design and it was just easier for me to start from scratch using OpenScad which is the only thing I know how to work with.

Anyway, 3 off E3D Titan extruders fitted to a custom mount, suspended above the centre of the bed on a pulley/counter weight arrangement. 905 gms in total but effectively weightless as far as the X carriage is concerned. There is of course some drag from the wires and tubes but nothing compare to lugging the mass of 3 extruders around. The extruder mount is tied to the X carriage with cord such that the Bowden tubes are always under a slight compression and won't get pulled out.

Bowden tube length is 300mm which I don't think is bad for a print area a shade under 400mm x 400mm on a frame which is 600mm x 600mm.