Filament grinding

To recap. I still haven’t solved this issue.

  • Without the enclosure prints still fail.

  • If I point a rather powerful fan at the printer during prints I succeed.

So, this is heat creep isn’t it?

Since the addition of extra cooling allows your prints to finish successfully it sounds pretty definitively like this is a heat creep issue to me.

Suggestions? Could the heat sink fan be failing? Or is there a suggested upgrade to a higher quality fan?

Working with PLA it honestly could just be an ambient temperature issue, PLA likes quite a bit of cooling and I’m not sure about where you are but it is still pretty dang warm out here and if the air surrounding the printer is especially warm it just won’t be as effective at cooling the nozzle. I know it’s already been mentioned by someone else but this could be a matter of gunk in the fan gumming up the rotation but aside from that I haven’t encountered them loosing power over time, from my experience it has been a pretty pass or fail sort of situation. I don’t personally have any suggestions for a better fan to use if you did decide to modify the printer, but there might be other people who have posted options here on the forums. If it was me I would probably just stick with the desk fan because it is easy and you might not end up needing the upgrade later on in the year once things start to cool down, but that might just be me always liking the easiest solution the best :laughing:

What tool head are you using?

Noctua has a 40x40x10mm (NF-A4x10 5V) fan that has a slightly higher output than the stock fan and a 40x40x20mm fan (NF-A4x20 5V) that is definitely higher output than the stock fan (but will require 10mm longer screws).

The manufacturer of the stock fan has higher output models but a minimum order quantity of hundreds of units!

My personal solution started with a 120mm, 120v AC muffin fan I pointed in the general direction of the tool head and that worked if I remembered to turn it off for the first few layers and turn it on for the rest (and there were failures because operator wasn’t always on the ball :slight_smile: ).

I have since added two 80mm 12V fans mounted on the frame of my TAZ 6 controlled by a Raspberry Pi (hat) synchronized to the speed of the cooling fans set by the gcode (M106, M107). My TAZ 6 is also connected to that Raspberry Pi running OctoPi / OctoPrint. This solved the “unreliable operator” problem :laughing: .

BTW, I’m still using the stock fan but I have the Noctua fans on the shelf if I need them.

Upgrade your print head to the Aerostruder and be sure to purchase and put on the silicone soc. I’m able to do 12 hour prints with PLA with this print head (and this is with a .2mm nozzle) Filament is really easy to load with the aerostruder print head too. The stock printhead would often fail in about 2 hours with PLA. I can’t see a reason to ever have to go back to the stock print head.

Yeah…I was just hoping not to have to use the fan.

I got the printer in May of last year. And I printed pretty much non stop for the first year. It just feels like I would have run into this problem last summer as well. The temperature was about the same.

I’ll investigate replacing the fan just to see what happens.

Stock.

Noctua has a 40x40x10mm (NF-A4x10 5V) fan that has a slightly higher output than the stock fan and a 40x40x20mm fan (NF-A4x20 5V) that is definitely higher output than the stock fan (but will require 10mm longer screws).

The manufacturer of the stock fan has higher output models but a minimum order quantity of hundreds of units!

My personal solution started with a 120mm, 120v AC muffin fan I pointed in the general direction of the tool head and that worked if I remembered to turn it off for the first few layers and turn it on for the rest (and there were failures because operator wasn’t always on the ball > :slight_smile: > ).

I have since added two 80mm 12V fans mounted on the frame of my TAZ 6 controlled by a Raspberry Pi (hat) synchronized to the speed of the cooling fans set by the gcode (M106, M107). My TAZ 6 is also connected to that Raspberry Pi running OctoPi / OctoPrint. This solved the “unreliable operator” problem > :laughing: > .

BTW, I’m still using the stock fan but I have the Noctua fans on the shelf if I need them.

This is very interesting, So you never tried one of the nocta’s to replace the heatsink fan?

As for the octoprint, I did just get a Raspberry Pi, but I haven’t gotten around to setting it up because of this overall problem :slight_smile:

I’m not against upgrades…but I was printing fine for more then a year, and I want to fully understand what is going on here.

Have you cleaned the hobbed bolt?

Are you using default Cura profiles from LB? If not revert back to the original profiles and try printing.

I’ve read other machines have benefited from a little cooking oil on the filament (PLA) to help it pass any buildup in the heatbreak or cold side… search thingiverse for filament oiler. Never tried this so YMMV.

Cleaned the hobbed bolt with a brush and then a dental pick (gently) to make sure there wasn’t anything stuck in there.

I had the same thought about settings, So I reverted to default, had a worry that it wasn’t properly resetting to default, then discovered that there was a newer version of LB cura available and have been using that.

Filiment oilers? Yeah I had read about those, I actually printed one but I don’t have any mineral oil. Keep meaning to order/go buy some. But I think we’ve decided that isn’t the issue.

Canola oil seems to be popular also…

Yeah I’m just saying that there’s been sooooo much discussion here on these forums about failed prints with PLA filament. Let’s understand that PLA has one of the lowest melting points of the most common filaments and is one of the cheapest filaments price wise. Unfortunately it’s melting point is just low enough for all the heat sources to cause it to fail. Ambient heat, heat rise from the print bed, extrusion speed, extraction and retraction moving the filament up and down as it softens, all contribute to the perfect storm.

I actually purchased a 11cfm fan to replace the stock fan but before I was able to swap it out I eased up the tension on my idler screw on the hob bolt and started making it through the prints. However most my prints were under 2 hours, And now my prints are fully printing. Back and forth between various print heads, height of summer now and again experiencing failed PLA prints, EXCEPT with the new aerostruder print head. So I theorize that the stock print head and the dual v3 printhead is just not designed well enough to prevent heat creep for PLA. There needs to be a better design to keep the PLA filament cool. The aerostruder print head may have been designed to cool the filament quicker to make longer and straighter bridges across gaps but the design also seems to not cause heat creep on PLA.

I’ve thought about running the PLA filament through some ice gel tunnel somewhere between the print head feeder and spool of filament. I suspect the room temperature still isn’t cool enough at 76 degrees.

Oh I see what you mean. That just might be it, but wouldn’t I also have run into this problem last summer? I got my Taz 6 in May 2017, and printed like a fiend for the first few months without issue.

And remember when I said prints were succeeding with the fan pointed at the printer? I just tested that again, and had a failure at about an hour in. Same story, the filament gets ground…

I just had a thought. I have a box of PETG and T-glase. I know these both require signifigantly higher temperatures, wouldn’t trying one of these be a good test of my issue(s)?

Why don’t you just chalk it up as heat creep is a problem with PLA and move on? No sense on mentioning that you should print with PetG or T-glase and all your problems will be solved cause you know damn well your problem will be solved only not with one type of filament, PLA. Have you read the this thread on the subject? https://forum.lulzbot.com/t/dual-v3-a-review-wait-for-v4-heat-creep-still-a-problem/6017/3 He did some thoroughough research and testing. Please compare notes and see what you’ve tried and what you’ve missed. Build up a database so we can all benefit from it. Get the temperature stats of the days it was printing correctly, including humidity and atmospheric pressure. Also, get the quality control reports from the manufacturers of the filaments, both successful prints and failed ones, I hope you have the batch lot numbers as that will greatly assist the manufacturer in getting this critical information you’ll need. Maybe you’ll uncover that they’re not mixing in enough hardner when they’re making a spool, a cutback that save the company thousands of dollars but unknowingly compromised the quality of the filament. Also, record the co-efficient of friction on all the moving parts of your printer, I can’t imagine a fan slowing down a bit from particles of dust landing on it over time. Let us know when you figured it out Sherlock cause many of have tried to help but we don’t have all the answers to cause and effect of things. Have you even tried changing out the print head cooling fan as many have mentioned? A desktop fan blowing in the direction doesn’t work. 3D printing utilizes very precise and efficient methods, the cooling fan can be one of them. We tried! Ask Alexa, Siri, Cortona or Google Assistant for the answer. Good Luck in this journey Pal!

The difference between last summer and current, it could be material build-up inside the hotend. Filament moving through the hotend could leave remnants of PLA along the sidewalls. Jams or failed prints where the hotend is at temp for lengthy periods can carbonize the PLA and potentially reduce the travel path.

I’ve always thought that the heat creep is mostly from the heater block. The suspect component could be the thin walled heat break which insulates the cold side of the hotend. Reducing / minimizing retraction during prints will help keep the tube cool. Ideally retraction should create just enough of a vacuum to reduce filament ooze during travel. To pass possible constrictions, increasing pressure from the idler should help the hobbed bolt/gear grip and push filament through. Lubrication could also help.

Of course there is a point where too much pressure from the idler can impede the hobbed gear/bolt from functioning efficiently. LB has a picture somewhere of “optimal” teeth marks on the filament… but the teeth marks and printing is really the only way to tell if you’ve got good pressure on the idler.

I had the same issue very recently. I switched to PETG and the issue stopped completely. My room (very small garage) would routinely get to 90F and I didn’t really notice the issue with PLA until it got that hot and started failing. I tested yesterday (three hour print) very early morning before the outside temps got above 80 and the PLA did fine. Might not be your issue but the heat certainly seemed to cause my issues.

I’ve been having this ‘filament grinding’ problem too; it started right after I updated the EEPROM firmware and Cura last week. So far every print I’ve tried to make fails at about z= 2 to 3 mm. When I remove the filament it looks exactly like the pictures posted by JRAlvarez.

There is a lot of discussion in this thread and some good advice, but I’d really like to hear the Lulzbot people coming up with some answers.

I am running into a similar issue with filament grinding. One thing I’m noticing is that the PLA I’m using has a rougher finish texture compared to similar filament in a different color. The filament in question is Inland 2.85mm blue PLA purchased from Microcenter. An Inland white PLA also from Microcenter has a much smoother texture.

A new spool of blue will print fine but as the spool gets used up, I see progressively more failures.

My theory is that as the filament wants to curl more and more inside the poly tube guide, because of the rougher texture, it generates more friction causing the grinding. I just tried pulling the filament through the poly tube and it definitely had higher resistance than I was expecting.

Another thing I’ve noticed is the 3D printed filament guide in the idler assembly (original extruder) is wearing out. It has a definite notch on the side where the filament pushes against. I’m suspecting this may also be contributing to the issue. I’m going to reinforce it with some brass or aluminum sheet to see if it helps.

Hope these observations helps others.

Since Inland is rebranded eSUN filament and eSUN might give the lower quality stuff to white label brands, I’d suggest buying eSUN brand PLA/PLA+ to see if the problem persists. INTSERVO is now eSUN’s one and only official US distributor (Aleph Objects sold out their US stock of eSUN filament two years ago).