A-Frame Spool Holders - Fundamentally Flawed?

[Dave]

Can you guys give me a quick sanity check please?

So out of the box, one problematic aspect of the stock FT-5 (R1) is the filament delivery.

The factory PVC pipe stub has issues with the filament drooping off the roll and tangling between the roll and the frame. It is also problematic in that the protruding socket cap screws securing the mount to the frame catch the webs of the plastic spools that FolgerTech sells, causing everything to bind up. Swapping those to button-head screws would solve the webbed spool catching, but still wouldn't keep the filament from escaping the sides of the spool.

There are quite a few alternatives for filament management - multi-piece guides to direct the filament over the top of the 2020 frame, reverse bowden setups, standard bowden setups, top-mounted spool holders, remote mounted spool holders, wall mounted spool holders, vertical posts that pull the filament off sideways, the FT-5 R2's top-rear-center mount, etc, etc.

One of the most popular alternative designs on Thingiverse is the triangular "A-Frame" style, which uses two triangular uprights on the top 2020 extrusions, with either a piece of conduit, threaded rod, or PVC pipe between them to suspend one or more spools of filament. These optionally have a separate guide arm that sticks out from the top front extrusion, to ensure that the filament is always pulled from approximately the centerline of the spool, not off the side, regardless of the print head motion. I'm talking about any of the dozens of designs that look generally like this one:

I thought these looked nice, so I picked one I liked and printed a set. They work fantastic for small prints, but I've had several failures with larger prints (larger in terms of X-Y dimensions), and after seeing the failure mode, I can't see how this design would ever really work well without being constantly monitored...

The theory makes sense. Locating the filament at the center of the build platform reduces the amount of filament that needs to be pulled from the spool when the print carriage moves to an edge or corner of the build platform. Or perhaps more accurately, it reduces the maximum amount of filament that is pulled, even if the average ends up being close to the same. That's the theory.

But when printing a large part in practice, a carriage movement to a corner will pull off a fair amount of filament, and a traversal through or across the build platform center point (or the point immediately below either the filament guide or where the filament exits the spool) will then cause that extra filament to bunch up. It's not going to rewind the spool, so that filament has to go... somewhere. Simply homing the printer and then moving to start a print in the center takes the filament from its maximum reach to its minimum.

On my printer, any significant lengthwise movement on the X-axis when the filament was unloaded (from a tension standpoint) and pushed like that would cause that extra length of filament to jump off the side of the spool. From there, it would bind up on the support pipe when pulled.

Maybe it was worse because my spools are relatively full, or maybe they have smaller flanges than others... but aside from some small item prints, I just couldn't walk away from the printer with this spool holder without coming back to a mess of filament hanging from the pipe beside the spool. I tried with and without a filament guide loop under the spool, and both had the same eventual result.

Did any of you encounter the same issue? I'll probably move to a reverse bowden setup to fix this, but I'm a bit bummed that this didn't work out, and surprised at how popular these are on Thingiverse if they are in fact fundamentally flawed in their operation...

[csorrows]

I went with the stock spool mount and reverse bowden on my original FT-5. I have yet to complete my R2 (using to complete the manual) so I don't know how that one works in real life. More than likely, they will both end up reverse bowden.

[JoeKilday]

I had the same problem, and stepping through the processes could see the flaw in the A Frame.
My solution is the same as csorrows from what I understand he described.

I use the standard holder with an Direct Drive E3D Titan + v6 which comes with a long PTFE tube and nipples, and build a nipple arm.

The PTFE tube keeps the filament from getting caught on things, and because it's curled in the same direction the filament is wound on the spool, it also works like the cable carrier.

This system doesn't unspool the roll when traversing the hot bed end-spaces.
Retraction is the only thing putting filament back the other direction.
Extrusion is the only thing pulling filament off the roll.
Slight tube friction keeps everything nice and tight in all the right places.

edit: lemme take a picture real quick...

[ChrisHenry]

While printing the triangle supports, I had similar issues with the stock R2 filament holder, where the spool would unspool if the printer reversed direction on a long axis move, essentially needing to "return" a length of filament, which was especially noticeable on a new spool.

To solve this, I ended up using a custom part, some PVC, and a short length of drip irrigation tubing. The net effect of this is that when the printer "comes back" after a large move, there's just enough friction in the tubing that the filament doesn't go back to the spool, but bends in space while the print head moves around and consumes it. I do get some rubbing of filament on the X axis cable chain, but it hasn't really been a big deal.

I've used this setup on mostly spent spools as well as brand new ones. Works great. Not the most attractive setup, but it works great, and doesn't require the triangle supports. You could also do a PVC pipe spool holder if you wanted as well.

Here's the STL file if anyone's interested in printing one for themselves...
Half inch PVC mount to 2020.stl (4.4 MB)

Photos:

[dvdsnyd]

I never liked the stock filament holder. The idea that the center of the spool just rests on static cantilevered beam leaves a lot to be desired. When Full a spool has quite a bit of momentum when moving and inertia when at rest. Someone(Can't find it right now) did a video and showed that imperfections in prints can be attributed back to filament delivery. He showed that adding bearings significantly helped. I included a picture of my setup, which uses a threaded rod and two printed cones and some bearings that effectively capture the spool. This allows the spool to be pulled with little force. The tension required to pull can even be adjusted based on how tight the wing nut is tightened down. it is finished with a reverse Bowden setup that helps guide the filament to the extruder and prevents the filament from coming off and tangling itself. This is the best picture I have on my phone at the moment. I can try to add more if people are interested.

I originally had the holder mounted vertically off the side of the machine. However, I am upgrading my hatch, and have aspirations to enclose the printer sometime soon. Moving the holder to a vertical position reduces the footprint of the machine, and doesn't seem to have affected it's functionality one bit. I actually think it's easier to use and looks nicer than flailing off the side.

[Ron]

I have something similar to yours only the axis runs horizontal off the left side of the printer.
It uses ball bearings and has worked quite nicely so far.
I can still add side covers and they won't get in the way.

I use a PTFE tube to guide the filament from the reel.
I have done that with both reverse and true Bowden setups and it works nicely either way.
With the stock reel hangar I used to experience filament tangles all the time.

I don't have any photos handy at the moment, sorry.
But if anyone wants to see them I can scrounge some up this evening.

I like your compressed air plumbing BTW (at least that's what it looks like).
I did something like that in our new garage.
The compressor is in the detached (new) garage and a line runs underground to the attached (old) one.
It is nice not to have to listen to the compressor all the time.

[dvdsnyd]

I like your compressed air plumbing BTW (at least that's what it looks like).I did something like that in our new garage.The compressor is in the detached (new) garage and a line runs underground to the attached (old) one.It is nice not to have to listen to the compressor all the time.

This is my basement. The plumbing you are referring to is for my sprinkler system
Would be nicer if it was for compressed air sometimes!

Dave

[Ron]

Aha, well it still looks nice.
I had wanted to build a new garage for many years, so when I finally did I went all out.
All the air plumbing is hidden behind sheet rock or underground.
Copper in the walls, but Polyflow in a conduit underground (for serviceability).

As for the spool holders, it is worth mentioning that a little friction is needed to prevent the filament reels from unwinding all by themselves.
So I used less than perfect bearings (rubber sealed) and keep a pre-load on them to make some drag.
When I was running the stock reel holder as-is I literally had to hand feed the filament to keep it from snagging.
So a decent reel holder was a top priority and an early print for me.

I had looked at A-frame holders from time to time but have other plans for the top of the machine including a raised cover someday.

[Dave]

To solve this, I ended up using a custom part, some PVC, and a short length of drip irrigation tubing. The net effect of this is that when the printer "comes back" after a large move, there's just enough friction in the tubing that the filament doesn't go back to the spool, but bends in space while the print head moves around and consumes it. I do get some rubbing of filament on the X axis cable chain, but it hasn't really been a big deal.

Thanks @ChrisHenry. That looks like a good solution - one that might actually be adaptable to the A-Frame spool holders I already printed. The irrigation tubing section as a friction-based backfeed preventer is a great idea. I suspect another aspect that makes your system work better is the height - your raised filament guide ultimately feeds out quite a bit higher than one pulling straight down off of the front of the spool. So that way, just with some simple geometry, your filament will bunch up much less than one with a lower outlet, and as an added benefit, it has more space for that bunch-up to go. The solution for my holder then would be to either print a second A-frame to use as an "up and over" guide, or to use a PVC structure such as what you have constructed here. Either way, looks like an interesting solution to the problem, so thanks for posting your photos!

[Dave]

I went with the stock spool mount and reverse bowden on my original FT-5. I have yet to complete my R2 (using to complete the manual) so I don't know how that one works in real life. More than likely, they will both end up reverse bowden.

I had the same problem, and stepping through the processes could see the flaw in the A Frame.My solution is the same as csorrows from what I understand he described.

I use the standard holder with an Direct Drive E3D Titan + v6 which comes with a long PTFE tube and nipples, and build a nipple arm.

I included a picture of my setup, which uses a threaded rod and two printed cones and some bearings that effectively capture the spool. This allows the spool to be pulled with little force. The tension required to pull can even be adjusted based on how tight the wing nut is tightened down. it is finished with a reverse Bowden setup that helps guide the filament to the extruder and prevents the filament from coming off and tangling itself.

I have something similar to yours only the axis runs horizontal off the left side of the printer.It uses ball bearings and has worked quite nicely so far.I can still add side covers and they won't get in the way.

I use a PTFE tube to guide the filament from the reel.I have done that with both reverse and true Bowden setups and it works nicely either way.With the stock reel hangar I used to experience filament tangles all the time.

@csorrows, @JoeKilday, @dvdsnyd, @Ron -
Sounds like there is a lot of support (and success reports) for the reverse Bowden setup.

One quick question - Since the tubing isn't between the extruder and the hotend, I'm assuming the filament path wouldn't need to be tight and "fully" constrained? Would there be a friction reduction benefit to using 3mm PTFE tubing for 1.75mm filament, possibly improving the retraction characteristics in the process?

[Ron]

One quick question - Since the tubing isn't between the extruder and the hotend, I'm assuming the filament path wouldn't need to be tight and "fully" constrained? Would there be a friction reduction benefit to using 3mm PTFE tubing for 1.75mm filament, possibly improving the retraction characteristics in the process?

In a word, yes.
I used 1/4" Polyflow tubing for mine initially, mostly because that's what I happened to have on hand.
The filament slipped around inside it pretty easily but it was not as slippery as PTFE would have been.
For reverse Bowden a tight fit is not critical at all.

[csorrows]

One quick question - Since the tubing isn't between the extruder and the hotend, I'm assuming the filament path wouldn't need to be tight and "fully" constrained? Would there be a friction reduction benefit to using 3mm PTFE tubing for 1.75mm filament, possibly improving the retraction characteristics in the process?

I am using 4mm ID 6mm OD PTFE tubing. PTFE provides very low friction and the larger size makes it a lot less prone to pushing on the wall. You will see this size used on most reverse-bowden setups. I works wonderfully.

[Dave]

This is the best picture I have on my phone at the moment. I can try to add more if people are interested.

@dvdsnyd -I didn't notice it at first, but I now see that you also eliminated both of the cable chains. Was there a particular reason you went this route?

[dvdsnyd]

One quick question - Since the tubing isn't between the extruder and the hotend, I'm assuming the filament path wouldn't need to be tight and "fully" constrained? Would there be a friction reduction benefit to using 3mm PTFE tubing for 1.75mm filament, possibly improving the retraction characteristics in the process?

I went with standard 1.75mm ptfe tubing with mine. Haven't had any issues. For a reverse Bowden setup, I don't think it much matters. However, with true Bowden style extruder, I imagine that the larger the difference between the filament diameter and the ID of the Bowden tube could increase complications with trying to run flexible filaments.

[dvdsnyd]

@dvdsnyd -I didn't notice it at first, but I now see that you also eliminated both of the cable chains. Was there a particular reason you went this route?

@Dave
Yes, it was very much deliberate.
When I purchased my kit late 2016, one of the biggest gripes I read a bout was the hassle of the cable chain and issues with breaking wires. I had originally thought it can't be that bad, but after seeing how TINY the included chain actually was, and that the individual links didn't open up. I decided to forgo installing it, since I knew I would be upgrading and doing various modifications to the printer. I didn't want to spend a bunch of time threading wires through the chain over and over again.

I have had good success with my set up. It's a little messy in those pictures, but I am currently upgrading to 24V PSU and making some modifications to the hatch. I'm looking to address the wiring mess too. Overall, I am happy that I went without the cable chain. Although, I understand that the newer kits have better chain, so it might be a better option now.

Dave

[csorrows]

I went with standard 1.75mm ptfe tubing with mine. Haven't had any issues. For a reverse Bowden setup, I don't think it much matters. However, with true Bowden style extruder, I imagine that the larger the difference between the filament diameter and the ID of the Bowden tube could increase complications with trying to run flexible filaments.

With a standard bowden setup you MUST go with the smallest ID PTFE possible. Backlash is a real issue with bowden setups. Even with PLA, it's still like pushing a wet noodle, less wet, but it still flexes a lot.

[Dave]

@dvdsnyd - Makes perfect sense. I didn't add enough wires when I assembled my R1, so I hit the same issues. I have a flexible 4-conductor wire zip-tied to the side of my cable chain, but found that the clearance is actually very tight where the Y-axis cable chain is routed. It's right up against the extrusion on one side, and anything attached to the left side would be hit by the X carriage when it moves to the endstop. So I have my additional wiring tied to the first set of the cable chain, but then in a carefully positioned free loop for the Y-axis movement to avoid snagging anything. It's far from ideal.

I don't think I would take apart what I have, but if I ever needed to, it makes a lot of sense to just make the wiring follow the same path as the bowden tube. After all, they follow the same movements, and could benefit from each other for mechanical support.

[diarx]

@Dave

The second photo in your post is actually my printer, I knocked up that filament holder fairly quickly as a stop gap solution, for it to work you really need that shorter A frame holder, not the one in your first photo, and you will need to align the A frame so that it places the edges of your spool as close to the filament guide as possible. Of course you're going to have to realign the A frame for different sized spools if you're using many different types, and some smaller spools may not work. I generally use standard size spools from the same place so it's a non issue for me.

The only problem I run into with my setup is I've upgraded to a 15mm x 15mm open cable chain, so sometimes the filament can catch a little on the chain because of the guide, but it never really causes any major issues. Unfortunately the enclosure I built for the printer is designed around these smaller A frame mounts, so I can't really accomodate a taller A frame setup without a major design rethink. Was also a rush job to get the enclosure built whilst I still had access to a workshop so the design is what I'm stuck with. Will eventually find a neat way to mount the spool externally to the enclosure I guess. Seem to have way too much going on lately to put loads of time into the printer, so I just keep it in a less than optimal but working condition.