Reduce strain on filament

Hi, in the search for the holy grail in 3DP, I stumbled upon my filament feed ... whenever the extruder is in position #1 or #3, the pulling forces on the extruder seem to be a bit high (I see differences in the outcome of surface quality), whereas in position #2 it seems fine, because while being in #1 and #3, it pre-pulls some filament and lives from that in/during #2. (See picture). I am using the Greg's Wade extruder. I thought about converting to Bowden, but to be honest, I'm not very motivated in doing that.


Of course I've already printed 608 bearing stuffed spool holders and it rolls off nicely and smooth, but the bearings can't help the weight of the spool, it still has to be moved and still requires quite some pull force which has to come from the movement of the extruder block.

So I thought I needed a solution. I had two solutions in mind and I'd like to hear your opinion about those.

The first solution that came into my mind was a software solution: whenever the extruder moves to #1 or #3, the machine makes a pre-pull filament move with no extrusion and then continues to extrude from the edges of the bed (#1 and #3) towards inside the bed (#2). In a rough estimation, this would eliminate any pull while extruding. This would require to modify slicing software though. Or does such a feature already exist and I just don't know? However, this constantly puts additional stress on the mechanical parts which requires sooner re-calibration, and possibly, the printing time would increase.

The second solution in my head was a hardware solution: while coming off the spool and being pulled or not, the filament bends more or less. With that discovery, an optical sensor could determine whether the filament is over a certain level of tension or not and then give a signal to an Arduino. The Arduino then would drive a little belt which moves the spool a little forward until the sensor says "ok, thank you, it's enough". We all have spare Nema17s, GT2 pulleys and belts laying around, that would make good use of that, wouldn't it?


Do any of such projects (or in the case of slicing software: features) already exist? If so, would you mind sharing a link? Or what other solutions are there?
Thanks in advance
Michael

On my CoreXY printers, I have the spool on top of the printer on the rear left corner at a 45 degree angle. The spool just sits on a printed spacer that then has an 8mm shaft (left over piece from cutting my rods) that's supported by two printed legs. Because the printed spacer in the hub of my spool puts the 8mm shaft in the dead center, the spool just freely spins and doesn't require any excessive force to pull the filament off.



And on this, I wanted a quick change, so the top of each side has a cap that just stops the 8mm shaft from coming out the sides and its really simple to change a spool. Took a few iterations, but I haven't found any major flaw to it.

Edited 2 time(s). Last edit at 11/13/2017 03:40PM by PDBeal.

A couple of quick comments;

The amount of force needed to pull filament off the spool seems to me to be an order of magnitude less than what's needed to push it through the hot-end. In the end, if your extruder doesn't have excess power to pull the filament from the spool *and* push it through the hot-end simultaneously, you're doomed, no matter how you arrange things.

An easy mitigation is to have the spool axle hanging from the ceiling by long cords (e.g [www.thingiverse.com]), so that the spool can relatively easily follow the motion of the extruder. (The same issue will also apply in the Y axis, if you have a printer that moves the extruder in the Y axis (e.g. typical Cartesian printer).

I understand what you're saying about the force required to pull the filament at positions 1 and 3, but can you point out specific print flaws that are related to that?

If you really want to solve that problem, build in a second X axis that runs parallel to the first and moves the filament spool with the extruder. Of course, a full spool, being a lot heavier than the extruder, is going to require a much bigger motor to move at the same acceleration/jerk as the extruder carriage...

A lot of 3D printing "innovations" are solutions to non problems. Dual drive extruders, for example, while mechanically interesting, aren't any better at squirting filament than single drive extruders with properly adjusted pinch roller pressure. If a printer is so marginally designed that the pull required to move the filament is causing print quality problems, the answer is to get more torque into the X axis. Anything that operates at the absolute margin of its performance is a disaster waiting to happen. Would you drive your car with the engine rpm red-lined all the time?

In early planning stages of UMMD I considered hanging the filament spool inside the enclosure, from the top cover. The hanger would swivel on ball bearings so the spool would turn to follow the extruder which moves in X and Y. I ultimately ditched that idea because it would require the top cover to be too high above the extruder carriage, but if you have a lot of vertical space, it would be a work nicely. The most common way of dealing with the "problem" you're trying to fix is not to force the filament to pass through the ring at the center of the frame. In many printer designs the filament is fed in a big loop through a piece of teflon tubing that will flex and follow the extruder carriage around. Enclosed printers that put the filament spool on top of the enclosure often feed through a slot in the top cover that runs parallel to the X axis. Other solutions include simply mounting the spool on a wall, well above the printer.

Edited 1 time(s). Last edit at 11/13/2017 10:16PM by the_digital_dentist.

---

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

Provided the extruder is on the carriage, the issue is not with the extruder pulling the filament, but fast movement of the carriage that has to pull it violently. It is the whole mass of the spool that has to rotate, not really the friction of its bearings causing problem.
So the filament path must be as natural as possible with no or minimum changes of direction, no filament guide if possible or one made with a short section of 4mm PTFE tube attached to something elastic like a simple piece of rubber band, or made with flexible filament (I made both and they work perfectly). Elastic to absorb the shock and PTFE for its low friction.
Thingiverse is a fun place to visit to see all the amazingly complex filament guides that fail to follow these simple rules

Edited 1 time(s). Last edit at 11/14/2017 04:22AM by MKSA.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Thank you, dear reprappers, for your valuable input! As regards my spool, it is currently located behind the printer, roughly about on the same height than the upper bar of the frame where the filament guide is located at; I haven't mentioned that in my first post.

@the_digital_dentist: On the side where there is strain on the filament, the perimeters sometimes show anomalies and if I set the speed down, I think I can see a very little lift on the X-carriage.

@MKSA: You're absolutely right, it is the movement of the carriage which actually pulls, I think I can neglect the extruder motor as a disrupter. And yes, I also have the impression, that my filament guide is not helping very much, I only wanted to get a defined way over my frame and not scrape off filament over the frame's edge, but the PTFE tube is a good idea, I will think about it! In the end it's movement against the spool's weight. So more torque on the X axis would possibly help, yes, that could be one route. Will also keep that in my head.

In fact, I do have a LACK enclosure with two LACK tables stacked (no doors yet, but that setup is already helping much with heat runaway prevention), and I have a spool dry box, so I could actually put the spool box on top of the lack table, pierce a hole through the box and the LACK table top and call it done. I understand from your comments and from @PDBeal 's picture that the filament-from-above solution rolls off much easier if there's nothing else in the way ("easier" compared to my current setup with the ring in the way). I would make an elongated hole in the direction of the x axis, so there will not be too much bending around the hole's edge.

That spool axle hanging from the ceiling by long cords seems to be a bit shaky for my case: if I mount it on the LACK table construct, I think it is a bad idea, because the table sometimes shakes a bit from the printer movements and those swing movements could add up. I also fear that this might start to swing back and forth and that it provides another backlash source.

So I think, I will try the Filament-from-above solution with an elongated hole and then compare the results.
Thanks again folks, I appreciate it!

Get rid of the filament guide. I have my roll set up as you have explained yours is now, and I don't have any issue. The filament just slides back and forth on the upper frame.

Beware also that the movement of the extruder relative to the filament source can cause other issues. On my PrintrBot, the PLA filament wore away the edge of the hole leading into the top of the aluminium extruder, eventually making it keyhole-shaped and causing jams. My fix for this was to drill out and tap the hole for M6, and put in a section of a broken stainless steel heat break.

Use a reverse-bowden setup.

The extruder should be what is pulling the filament, never the motion of the axes.
Otherwise you are bound to get artifacts on your part from trying to drag the spool when the x-axis moves.
Reverse bowden completely eliminates this, and it's a simple retrofit, keeping all the advantages of direct drive extruders.