Filament grinding at low-speed details with PLA

Share your successful printing settings with others here. If you have any question about the parameters or printing skills of some filament, please find here.
wermy
Posts: 23
Joined: Fri Jan 20, 2017 1:49 am

Filament grinding at low-speed details with PLA

Postby wermy » Sat Feb 11, 2017 6:25 pm

I'm not sure why this is happening on my N2 with PLA. I've tweaked all kinds of settings trying to fix this, including temperature, extrusion multiplier, extrusion width, printing speed... I've tried about 10 different combinations, but every single time, when my print gets to the thin structures in the picture below, which it slows down considerably for (and are taller than the rest of the print, so that's all that is left to print at that point), it just stops extruding and starts grinding filament. The rest of the print is perfect. WHY IS THIS HAPPENING? I'm going crazy! I'm using Simplify3D for slicing, by the way.

Last settings I tried (shown in the picture):
0.95 extrusion multiplier
0.4mm extrusion width
220 extruder temp
40mm base print speed

Any help would be greatly appreciated. Thanks!

IMG_1984.JPG

Jetguy
Posts: 1528
Joined: Tue Mar 22, 2016 1:40 am

Re: Filament grinding at low-speed details with PLA

Postby Jetguy » Sat Feb 11, 2017 7:02 pm

Adjust your extruder so grinding isn't possible. I would hope this would be obvious but apparently not, like a car spinning a tire in the mud, once you dig a hole, you aren't getting out without a push.
So, you have to be able to look at an extruder- see the drive gear, see the V -groove bearing and see the distance the filament is smashed into the drive gear. On the Raise 3D n series, that distance is fixed by how the motor is mounted to the aluminum bracket. You have to loosen the 4 screws holding the motor, with the filament removed, move the motor by hand pushing it closer to the V-groove bearing, then tighten the 4 motor screws and test.
The idea is that during a loading filament function, you should be able to grab the filament going into the top of the extruder with your hands and in doing so, cause the motor to skip steps but never, ever, grind the filament. Until you achieve that state, you are wasting your time trying to troubleshoot.

This is why many of us choose to upgrade the extruder in one of the following ways:
#1 the cheapest fix is put a sharper tooth drive gear on the motor. It's so cheap, I don't know why this isn't the way they come or dealers should supply these. https://www.3dp2go.com/38-teeth-steel-e ... 4veugqmcn1

#2 The next level is just flat out removing the entire stock metal extruder block and replacing with a Bondtech or other filament feeder section. This both reduces the moving mass and improves grip and reliability of the entire printer. Yes, more expensive but once you try it, you would not want to go back stock. Martin specifically worked and designed drop in extruder upgrades for the N series http://shop.bondtech.se/ec/

wermy
Posts: 23
Joined: Fri Jan 20, 2017 1:49 am

Re: Filament grinding at low-speed details with PLA

Postby wermy » Sat Feb 11, 2017 7:08 pm

Jetguy wrote:Adjust your extruder so grinding isn't possible. I would hope this would be obvious but apparently not, like a car spinning a tire in the mud, once you dig a hole, you aren't getting out without a push.
So, you have to be able to look at an extruder- see the drive gear, see the V -groove bearing and see the distance the filament is smashed into the drive gear. On the Raise 3D n series, that distance is fixed by how the motor is mounted to the aluminum bracket. You have to loosen the 4 screws holding the motor, with the filament removed, move the motor by hand pushing it closer to the V-groove bearing, then tighten the 4 motor screws and test.
The idea is that during a loading filament function, you should be able to grab the filament going into the top of the extruder with your hands and in doing so, cause the motor to skip steps but never, ever, grind the filament. Until you achieve that state, you are wasting your time trying to troubleshoot.

This is why many of us choose to upgrade the extruder in one of the following ways:
#1 the cheapest fix is put a sharper tooth drive gear on the motor. It's so cheap, I don't know why this isn't the way they come or dealers should supply these. https://www.3dp2go.com/38-teeth-steel-e ... 4veugqmcn1

#2 The next level is just flat out removing the entire stock metal extruder block and replacing with a Bondtech or other filament feeder section. This both reduces the moving mass and improves grip and reliability of the entire printer. Yes, more expensive but once you try it, you would not want to go back stock. Martin specifically worked and designed drop in extruder upgrades for the N series http://shop.bondtech.se/ec/


Thanks for the reply and for the links. I'll definitely be getting that drive gear, and maybe the bondtech at some point.

I'll try adjusting things as you said, but what really has me puzzled is why it's JUST on those very slow, taller structures that this happens? I watched/listened to the entire print, the drive gear didn't slip once. It only slips and starts grinding once it slows down for those tall parts. I would think it slowing down would actually make it less likely to grind filament, no?

Edit: according to Simplify3D it slows down to like 3mm/s during those parts:
Screenshot 2017-02-11 13.17.06.png

Jetguy
Posts: 1528
Joined: Tue Mar 22, 2016 1:40 am

Re: Filament grinding at low-speed details with PLA

Postby Jetguy » Sat Feb 11, 2017 7:21 pm

No, retraction and pressure can vary greatly and cause the system to generate more back pressure. A lot of this goes into hotend thoery of operation and that's not something I can easily post here and explain. Think of it like a test tube and you are jamming a tight fitting rod down into it. If you put a viscous liquid like say peanut butter into the tube, then when you jam the hard rod in, the peanut butter liquid tries to squeeze up into the space between the hard rod (fresh unheated filament) and the outside walls of the tube. However, the viscosity and action of the rod pushing down creates a shearing force pushing the liquid back down rather than it just squirting up around the incoming filament. That ratio and where that molten filament tries to push up to before being pulled back down changes with temperature, filament type and the speed and pressure at which we are extruding. Filament sitting and slowly extruding, the more time that heat begins to melt further and further up the hard filament. The hard filament becomes soft and then deforms and wedges against the walls under pressure. It's not fully melted so it's not sliding easily down the wall, but it's soft and deformed. This is how a cold end jam forms (or at least one theory and proven in practice).

So no, extruding really slow is not good for the system either and can be a cause.

wermy
Posts: 23
Joined: Fri Jan 20, 2017 1:49 am

Re: Filament grinding at low-speed details with PLA

Postby wermy » Sat Feb 11, 2017 7:29 pm

Jetguy wrote:No, retraction and pressure can vary greatly and cause the system to generate more back pressure. A lot of this goes into hotend thoery of operation and that's not something I can easily post here and explain. Think of it like a test tube and you are jamming a tight fitting rod down into it. If you put a viscous liquid like say peanut butter into the tube, then when you jam the hard rod in, the peanut butter liquid tries to squeeze up into the space between the hard rod (fresh unheated filament) and the outside walls of the tube. However, the viscosity and action of the rod pushing down creates a shearing force pushing the liquid back down rather than it just squirting up around the incoming filament. That ratio and where that molten filament tries to push up to before being pulled back down changes with temperature, filament type and the speed and pressure at which we are extruding. Filament sitting and slowly extruding, the more time that heat begins to melt further and further up the hard filament. The hard filament becomes soft and then deforms and wedges against the walls under pressure. It's not fully melted so it's not sliding easily down the wall, but it's soft and deformed. This is how a cold end jam forms (or at least one theory and proven in practice).

So no, extruding really slow is not good for the system either and can be a cause.


Ah! Thanks for the great explanation - I appreciate it. :)

Jetguy
Posts: 1528
Joined: Tue Mar 22, 2016 1:40 am

Re: Filament grinding at low-speed details with PLA

Postby Jetguy » Sat Feb 11, 2017 7:30 pm

What you are showing is what I call tower slowdown.

So you have an object with one or more tall areas where nothing else is being printed on that layer. S3D and many slicers have a minimum layer time in the cooldown section of settings (or similar wording) and the idea is that plastic takes time to solidify so if you layer down an area and then go right back over top and lay another hot layer, the whole things stays too hot to become solid and droops. Unfortunately, a fix that made it into common settings does (AKA defaults) does NOT work in all situations and can be bad. So again, that setting is often called minimum layer time. The way it works is gcode is just distances as targets. It also contains a speed value. But, when the slicer sees that all the gcode for that layer takes less than the minum time (speed and distance), then it slows down to make the layer take the minimum time set. You go the same distance- just longer time. The problem is, that means the extruder is going slower, the nozzle is also putting more heat into the spot it's moving over top of and thus minimum layer time doesn't work that great.

So the BETTER fix is to add features to the build area that make the layer take longer without slowing down.
Either print more than one small object at a time OR add tall "waste" objects like purge towers to again, simply take of time to make the slowdown never happen and your prints improve with minimal waste.

This is actually covered in the Simplify 3D quality guide and most of this again is universal to many slicers and 3D printers in general.
https://www.simplify3d.com/support/prin ... verheating

If you have already tried the 3 items above and you are still having trouble achieving sufficient cooling, there’s one more thing you can try. Create a copy of the part you are trying to print (Edit > Copy/Paste) or import a second object that can be printed at the same time. By printing two objects at once, you can provide more cooling time for each individual part. The hot nozzle will need to move to a different location on the build platform to print the second part, which provides a short relief for your first part to cool down. This is a simple, yet very effective strategy for fixing overheating problems.
Attachments
Over-Heating.jpg
Last edited by Jetguy on Sat Feb 11, 2017 7:44 pm, edited 1 time in total.

Jetguy
Posts: 1528
Joined: Tue Mar 22, 2016 1:40 am

Re: Filament grinding at low-speed details with PLA

Postby Jetguy » Sat Feb 11, 2017 7:43 pm

Here are the settings in S3D.
Attachments
S3D Prime Pillar Ooze Shield.jpg
S3D slowdown minimum layer time.jpg

Jetguy
Posts: 1528
Joined: Tue Mar 22, 2016 1:40 am

Re: Filament grinding at low-speed details with PLA

Postby Jetguy » Sat Feb 11, 2017 8:21 pm

I just want to make it clear, we need to solve the grip problem first.
Anything that creates greater pressure than the grip of the feeder causes it to grind. Once it grinds, because the stock system is a set distance to press the filament into the drive gear, you get zero pressure and the system cannot self recover without human intervention. If we go the opposite way and create near perfect grip that is stronger than the motor can overcome, a stepper motor can skip steps all day long and never damage anything. So worst case, once the jam or backpressure reduces, the motor can overcome and begin pushing filament again -aka self recovery.

So yes, by using the settings, might find a settings that don't stress the system or create conditions that reduce the tendency to jam and grind filament, but you are just masking a weakness.

Just saying, solve both problems as a top priority. Fix the mechanics now and better understand how they function, and change the settings to reduce instances of problems.

wermy
Posts: 23
Joined: Fri Jan 20, 2017 1:49 am

Re: Filament grinding at low-speed details with PLA

Postby wermy » Mon Feb 13, 2017 3:27 pm

@Jetguy thanks again, adding a second throwaway model did indeed help a lot. I do think I'll go ahead and put a bondtech extruder on there, though. I'm really not impressed nor happy with the stock extruder design. It looks sexy, sure, but it is SO much pickier than my lulzbot's. Hopefully this helps. I've already sunk ~$3k into this machine, what's another $175?

wermy
Posts: 23
Joined: Fri Jan 20, 2017 1:49 am

Re: Filament grinding at low-speed details with PLA

Postby wermy » Mon Feb 13, 2017 4:44 pm

@Jetguy:
Just ordered a bondtech. I see you have a custom firmware up here for it: viewtopic.php?f=4&t=181
I have a few questions about that:
I don't see it mentioned in bondtech's installation instructions, is it required or does it just implement the estep change in the firmware itself so you don't have to add it to the startup script?
If it is required, is the version linked to in that thread up-to-date?
Does it add any other nice features?

Thanks again for all your help, I see you jumping in and answering questions all over on these forums, and it's greatly appreciated. :)

Jetguy
Posts: 1528
Joined: Tue Mar 22, 2016 1:40 am

Re: Filament grinding at low-speed details with PLA

Postby Jetguy » Mon Feb 13, 2017 5:25 pm

On the Bondtech, you have choice of how you want to handle it because in the most simple of terms, it has a different steps per mm than the stock extruder.
You can do the bare minimum, keep the stock firmware, adjust the stepper driver Vref to 0.6V to prevent the motor from heating up, and just edit inside of Ideamaker the custom steps per mm and that gets added to each and every print gcode.
So downsides (subjective) of this method. The load and unload scripts will now move filament slower and the commanded distance values will be smaller as well. Might not matter to many folks, but I use the Mosaic palette with mine and precision loading of filament to exact mm is required to start a palette print. You just have to think about it, the front LCD panel is just sending gcode commands to the motion board. In this state, the firmware default steps per mm is assumed to be the effective value for those commands.
Or
You can also use Ideamaker and adjust the extrusion multiplier and get a similar effect as changing the steps per mm. Either way, it's all just added gcode commands to make the correction on a per print basis.
Or
You can use my older sources, or use new source and flash firmware that sets steps per mm.
Now in saying that, I simply turned on some heater safety functions as they existed in firmware and have been printing that way for a while. Once in a blue moon, I might accidently adjust the bed or heater temp manually in such a way it invokes a flaw in the safety code and then I have a reset the printer. Just saying, use my code "at risk". It would never harm your printer but if you often got caught in the safety code, then you might be frustrated.

Also, I know others were working on the firmware and I've been out of it for a while (mine works, I was happy and thus left it alone), so there are other ways of doing this and alternate firmware branches and methods folks worked on.


Return to “Printing Settings”

Who is online

Users browsing this forum: Google [Bot] and 1 guest