Avisos
Vaciar todo

MK3S filament sensor problem  

  RSS
johanrild
(@johanrild)
New Member
MK3S filament sensor problem

Hi, I'm having reliability problems with the new filament sensor. Luckily the problem is that it stays triggered so it keeps printing.

When I take off the sensor board I can freely move the blocker arm (fs-lever), but I notice that if I insert filament (through the correct opening) and slowly take it out, the blocking arm doesn't move back. If I gently touch the arm, then it moves back. While the magnets are powerful, they don't seem to have enough power at the end of the stroke (where it needs it).

To me this looks like a design flaw. The fs-lever should have the lip (the one resting against the ball) about 2mm longer out (I measured). I would do this myself, but the part is buried as deep inside the extruder assembly as you can get.

Does anyone know a remedy? Preferably without taking the part out and sanding all the mating surfaces.

I have attached a image showing what measurement I think is too short. (I know there may be good reasons to have the arm as short as it is, I don't have all the information)

Respondido : 10/03/2019 1:04 pm
marchyman
(@marchyman)
Eminent Member
Re: MK3S filament sensor problem

Loosen the pivot screw. Without the magnet installed the FS lever should flop back and forth with no binding. You might also want to check the housing and lever to make sure there are no bits of plastic swarf or other contaminants that could cause binding.

Respondido : 10/03/2019 7:31 pm
johanrild
(@johanrild)
New Member
Topic starter answered:
Re: MK3S filament sensor problem

Unless I'm mistaken, that screw is buried inside the extruder body. The lever feels very loose when I move it, but it binds when I move it extremely slow (as it does when the filament operates it). It might be fixed by loosing the screw, but I don't understand why they haven't designed it so the magnets are closer. Maybe it makes it harder to assemble.

Respondido : 10/03/2019 8:12 pm
steven.m25
(@steven-m25)
Eminent Member
Re: MK3S filament sensor problem

I was having issues with my filament sensor as well. Mine was not sticking or binding. And I have been printing many versions of this trying to get my IR sensor to work correctly, analyzing the level and placement and fit with the stainless steel ball.

I believe that the design (at least as I downloaded and assembled last night) is sound. I think my problem came from using some 3rd party filament (it was not blocking the IR). But if you got a set of parts printed make sure you use the Prusa supplied PETG. And if you can print it on a MK3, use their gcode.

There is very little play in the movement of the ball bearing. When a ball bearing moves it shifts that level 1.75mm so the tolerance is very tight. That little notch (the stairstep under the 6mm in your drawing) is the part that breaks the beam of IR. And from what I can see, it's created exactly where it needs to be.

If it's binding I'm guessing it's surface textures that are causing the problem. I don't think the magnets being closer or farther away will change anything with the design.

I'm not sure if you tried, but make sure you use their filament and their gcode and assemble.

Respondido : 11/03/2019 4:43 pm
steven.m25
(@steven-m25)
Eminent Member
Re: MK3S filament sensor problem

You could loosen the pivot screw, maybe. You could sand the top and bottom face of the lever if you want. I wouldn't mess with the tip of the lever that touches the ball bearing.

The ball bearing only moves 1.75mm and the tolerance is very tight there.

My thought is that Prusa makes a ton of these and I'm guaranteeing they don't do that much post processing on these parts to get them to work. They yank them off their printers and assemble. So if we have issues it's either our prints that we made or our assembly issues.

I reprinted my lever and extruder housing (everything that touches the lever) again with their filament and that fixed my issue. Also didn't screw down the pivot screw as much. That might be a touchy setting (how tight that needs to be).

Respondido : 11/03/2019 4:53 pm
johanrild
(@johanrild)
New Member
Topic starter answered:
Re: MK3S filament sensor problem

@steven.m25, I do realize that extending the arm as I suggested also entails shortening or moving the blocker lip. I do not agree that detecting 1.75mm movement requires tight tolerances. I do realize however that doing a redesign like that, way inside the extruder body, is too much work as I'm sure the first few iterations won't work. It will end with me sanding down all the mating surfaces on the FS-Lever, making sure the screw is loose and hope that it fixes all the problems.

I like the idea of using magnets as a spring, but unlike a spring, magnets loose force exponentially with distance (the force is inversely proportional to the distance squared). This is why I think a tiny distance change of only 2 mm would greatly increase the force and make the problem I have go away and make the design more robust. Using a spring would also fix the problem.

Respondido : 12/03/2019 5:21 am
steven.m25
(@steven-m25)
Eminent Member
Re: MK3S filament sensor problem

Yes, but I believe you are trying to fix a symptom not by fixing the problem but by changing the design to fix the symptom. I don't deny you are having a problem and you have a symptom.

And when I said tolerance maybe that's the wrong phrase. You were talking about the lever bring 2mm longer or the design changed. When I had my filament sensor problems, and I was checking the design myself, they have to have that arm exactly the size it is so that it just moves out of the way of the IR sensor and then blocks the sensor, there isn't that much difference between filament in and filament out.

I'm just saying that we had/have filament sensor issues. And I'm not seeing 100's of others saying the same thing. This is one of the big changes between MK3 and MK3S. And when I had my filament sensor problems, I started to blame the design, but I'm guessing Prusa doesn't do a lot of post processing and it doesn't look like there are a lot of people complaining about this. So it's most likely us.

My issue was printing the parts with different branded PETG. I took apart my extruder and used new parts printed with their gcode with their supplied filament (which sucked cause I had to put my printer together temporarily to get it to print so I could take it apart again). I was very careful and my problems were resolved.

If you want to redesign it, don't let me stop you. I'm sure there are tons of ways this could be redesigned and improved. But I think it functions properly with a good set of parts assembled correctly without a redesign. If you do redesign, please share on thingiverse or here so we can check it out.

Respondido : 12/03/2019 12:57 pm
johanrild
(@johanrild)
New Member
Topic starter answered:
Re: MK3S filament sensor problem

It turns out it wasn't as hard to get to the lever as I imagined. Since it was one of the first things attached to the assembly I thought I had to strip it down quite a lot to get to it. It turns out you can partially remove the motor and get a gap just big enough to get the lever out (using a ball head hex key as it wont fit strait). No need to cut any zip ties.
The fault was easy to find. The 3mm hole in the lever for the screw was a bit tight. Using a 3 mm drill fixed the problem. The screw was not tightened too tight by the way. With the looser interference (as in no interference) the magnet feels much stronger.

@steven.m25 I though I said I wasn't going to redesign it, I just said how I think the design could be improved. I know not to touch others design without having a lot of time, what looks like a easy fix from the outside often creates new unforeseen problems that has to be fixed. It turns out the magnet is strong enough with the part corrected anyways (I take back all the negative things I said about the design. It felt good to blame someone though).

Respondido : 12/03/2019 6:47 pm
steven.m25
(@steven-m25)
Eminent Member
Re: MK3S filament sensor problem

No worries, my friend. I'm glad you figured out your issue (and I figured out mine).

When I was trying to figure out my filament sensor problem I was convinced it was something wrong with the design, but turns out it was just me. 🙂

Respondido : 12/03/2019 7:46 pm
marchyman
(@marchyman)
Eminent Member
Re: MK3S filament sensor problem


The fault was easy to find. The 3mm hole in the lever for the screw was a bit tight. Using a 3 mm drill fixed the problem. The screw was not tightened too tight by the way.

Ahhh, binding on the shank of the screw, not being bound by the screw. Thanks for the update. Glad it was an easy fix for you.

Respondido : 13/03/2019 6:26 am
Jimbo70
(@jimbo70)
Estimable Member
RE: MK3S filament sensor problem

I have the same problem. Magnets are strong but not strong enough to push the lever back. I think the close tolerance makes the parts stick when they expand with the heat. I sprayed a very small amount of PTFE dry lube into the lever area after I removed the sensor. I was careful not to get the spay on the gears. The sensor is now working normally. I think this will prevent the sticking problem.

Respondido : 04/05/2019 5:10 pm
Compartir: