Page MenuHomeAleph Objects Inc

Banded 0.8 printing feedback
Closed, ResolvedPublic

Description

Feedback Checklist:
Have You Already:

  • Cleaned the Nozzle off/out?
  • Cleaned and/or Sanded the bed?
  • Leveled the X axis?
  • Restored Failsafe?
  • Updated your Firmware?
  • Updated your Version of Cura?
  • Can you reproduce your error reliably?

Error Description (this is the information that needs to make it into the code ticket)
Please explain your original intent and then the failed result.

Include this Information for R&D Use:

  • Machine Type
  • Toolhead Type
  • Firmware Version
  • Cura Version
  • Material Being Used
  • Estimated age of filament
  • Profile being used (please include any profile edits you made, no matter how small)
  • STL and/or Gcode (add as file, web link, or shared-j location)
  • Environmental Description (enclosed or open air?)
  • Pictures (to avoid data issues please remember to compress photos before posting)

Event Timeline

adam created this task.Oct 19 2018, 11:43 AM
adam triaged this task as High priority.
adam created this object with edit policy "LulzBot Hardware Products (Project)".
adam added a comment.Oct 19 2018, 11:48 AM

Supervisors: a new build a cura is available with profile updates for Banded Tiger. Please add each person testing a banded tool head to this ticket and instruct them to clear their cache and build a new version of 3.2.30 CuraLE.
Profiles ready for testing:
ABS (IC3D and CSL) high speed and standard
PLA (Verbatim and PolyLite) high speed and standard
PETg (IC3D) high speed and standard
NinjaFlex high speed and standard
PA-CF (colorFabb) high speed - apologies all, I don't have this material to hand out yet for general testing, but I should have some by end of day Monday.
Edits have also been made to the XSTRAND profile, but they haven't been pushed yet. Brent should have that updated soon as well as updating all the Dingy profiles.

adam changed the edit policy from "LulzBot Hardware Products (Project)" to "AO (Project)".Oct 19 2018, 12:30 PM
adam added a comment.Oct 19 2018, 12:38 PM

XSTRAND has been updated for Banded

nathan added a subscriber: nathan.Oct 22 2018, 8:10 AM

As noted already on the Dingy ticket, the print speed will need to be lowered to speeds like the Taz 6 Moarstruder, as the ones that are set currently are too high. Aside from that, I found that a few other settings allowed for a better surface quality, and that there is an issue with sharp corners that I have yet to really nail down. Here are some photos. These were done on a Mini with firmware version 1.1.9.23

Just the speed change:



(accidentally broke off the smokestack)

Speed change with "Outer Before Inner Walls" enabled:

There is a bit of a persistent issue with the layer height change, but that may just be a filament issue, as I also did some rocktopi with PETG (white is Innova-1800, although both were done using the same base standard profile). Here is one with the OBIW disabled with the speed change:


With OBIW enabled and speed change:

Hey Nathan, can you do a fresh profile pull (for reference: http://wiki.alephobjects.com/wiki/Cura_LE_Local_Build_Directions) and try these again? Also what material were you using? A bunch of things have been changed and I am hoping these new settings will provide better layer adhesion. I have been trying to slow down most of the profiles because the material and extrusion are much thicker but because they are being marketed as being a faster option I am getting some argument for faster speeds. I would like to know if these new settings fair any better for you and if not I can figure out the best fix. :D

Will do, bit busy over here today, so I might need to get back to you tomorrow, but it will be done.

Providing feedback from using the macro in a production sense:

The probe is hit or miss on how much material is picked up in the wipe. (currently wiping at 170)
If the probe went well it prints with no issues. The problems currently that are concerning for manufacturing is just how much manual wiping was required after 2-3 prints, there was enough material on the nozzle to cause probe failures. This is starting with a fresh wiper pad.

I will attempt a fresh pull to see if default profiles have a better result as time allows. In my opinion it will likely need a different material for the wiper pad due to how fibrous the current solution is. With the inevitable ooze and ABS buildup on the nozzle I think wiping will always cause more headaches than it solves with this nozzle style and material choice. Going hotter on wiping temps risks even more buildup on the nozzle as the glass transition at 105 c and the hotter we get the more material from the pad will want to transfer to the nozzle.

Greggb added a subscriber: Greggb.Oct 23 2018, 2:42 PM

First print with Banded Tiger and X Strand filament resulted in poor print quality:

Machine Type: Mini 2
Toolhead Type: banded tiger
Firmware Version 3.2.30
Cura Version: 
Material Being Used: X strand
Estimated age of filament: brand new
Profile being used: most recent
STL and/or Gcode: /home/aleph/shared-j/Documents/Creative/3D_Models/Made_By_LulzBot_Designs/General_Prints/Wrench/Wrench19mmx25mm_Closed_Small.blend
Environmental Description: open air
Pictures: /home/aleph/shared-j/Documents/Creative/Video/BandedTiger_Dingy/Images/First_Print/4X9A9863.JPG
karrad added a subscriber: karrad.Oct 23 2018, 2:50 PM

@Greggb We need a specific profile. Is that Standard or High Speed?

Sorry, this was high speed.

Steven added a subscriber: Steven.Oct 23 2018, 4:26 PM

@Greggb was that X strand straight out of the packaging or was it dried before use? We may want to order a filament drier for Lago. This is what is in use in R&D https://www.amazon.com/Dehydrator-Machine-BESTEK-Programmable-Temperature-x/dp/B01NAAL5XQ

@Steven this was straight out of the packaging, but it didn't have any of the signs of moisture Adam warned me to look for. No popping sounds or steam.

DRobertson added a comment.EditedOct 24 2018, 11:28 AM

Machine Type: Mini 2
Toolhead Type: Banded Tiger
Firmware Version: 1.1.9.17
Cura Version: 3.2.30
Material Being Used: PETg
Estimated age of filament: <2 months
Profile being used (please include any profile edits you made, no matter how small): Standard (0.4), High Speed (0.6), no changes made
STL and/or Gcode (add as file, web link, or shared-j location)


Environmental Description (enclosed or open air?): Open Air

PETg has been consistently successful for me using the Standard 0.4mm layer profile on Mini 2 and Cura LE 3.2.30. There are still some pokey parts on the top layer from the tool head hopping as it retracts though. Surface quality generally looks good with very few holes in the top layers and the walls, even on sharp angles.

The High Speed 0.6mm layer profile has been more problematic. Even with hopping for retractions the Y-axis is still getting bumped out of place on some prints and the prints themselves are weak and easy to break apart along layer lines.

I have completed 2 prints on the Polylite PLA profile at standard speed and found that there was a definitely stringing occurring, it might be good to adjust the retraction settings for this profile slightly. Otherwise it printed both of my parts just about perfectly without any adjustments.

Photos showing the stringing:

General photos showing prints:

Orias added a subscriber: Orias.Oct 25 2018, 12:09 PM
Lily added a subscriber: Lily.Oct 25 2018, 3:28 PM

First Prints with Banded Tiger using PolyLite PLA. The prints are not great, lots of stringing and the layers are messy (is this pretty standard for a larger size nozzle?). Image 1 is the first print, a rocktopus on the standard settings, normal size. Image 2 and 3 are both of a rocktopus that I maxed out my bed with (to test if the print was just too small for good details), this one was also printed at 210 instead of 220. I saw no difference in stringing with the temp change or size change.

Machine Type: Mini 2
Toolhead Type: Banded Tiger
Firmware Version: 3.2.30
Cura Version
Material Being Used: PolyLite PLA, Teal
Estimated age of filament: 1 week since I opened it brand new
Profile being used: Standard (0.4mm) PolyLite PLA Profile used for Image 1. I then changed the temp down from 220 to 210 in hopes that this would help the stringing and maxed out the print bed (second print, image 2 and 3).
STL and/or Gcode: https://www.lulzbot.com/content/meet-rocktopus
Environmental Description (enclosed or open air?): Open air, situated by a wall

@DRobertson For your high speed PETG print, pleas enable acceleration control, and change all values to 1000. We are making this change within the firmware, and would like to see if it helps your results.

@Galadriel @Lily Are those normal sized rocktopus, or scaled up? The banded will naturally have a harder time with those finer details like the fingers, we have also found the extra temperature is helpful for larger objects. (It prevents strip outs and weak layers)

Lily added a comment.Oct 26 2018, 9:05 AM

@karrad The first print is normal sized, per my description above, second print is maxed out (exact percentages below). Both show enormous amounts of stringing.
First print, image 1: 100%
Second print, image 2 and 3: 192.4%

@Lily Sounds good. I have one going as well, will post a pic here in 20 or so to compare

@karrad Mine was done at standard size because I was trying to see how far I could push the toolhead. Overall I was impressed, that bit of stringing I would normally just post process honestly but I know we have some customers who might see that as to much post processing so I thought I would mention it.

Not quite full build volume, but scaled up to ~150%: Banded, High Speed, 3.2.30, polylite PLA

I think one of the important parts here is to keep in mind here is to manage customer expectations. These tool heads are intended for strong large parts, fast. If there is a need for more accurate parts, it would be best to suggest going back to the standard extruder.

We can work on trying to help the stringing, but if it comes at a cost of weaker parts it is something we may not be able to eliminate completely.

Lily added a comment.Oct 26 2018, 9:32 AM

@karrad Right on. That makes sense!

Sorry I haven't responded yet! I have been fighting with stringing over here as well and I seem to have got Polymaker MUCH better but verbatim is still giving me a hard time. I will be posting updates today for polymaker, hopefully this will help next week :)

adam added a comment.Oct 26 2018, 11:47 AM

@anolen the bed is moving forward after my XSTRAND prints right away, not cooling at all. Can we leave it back until we cool the bed to a set temp?

Standard profile, 3.2.30, polylite PLA, scaled to ~150%

@adam, yeah, it is supposed to be staying back, that fix will be in the next update, thanks for letting me know.

Lily added a comment.Oct 26 2018, 3:08 PM

Next Test was on the Benchy STL. High Speed, temperatures changed to 210. Some stringiness and interesting Z-seams. The front of the Benchy looks kinds of wavy and talked to Nathan who said this was probably due to a draft around my printer. I'll be testing out the normal 220 temp and changing the z-seam next.

New profiles and firmware have been pushed. Please re-build master branch to get the latest and greatest.

adam added a comment.Oct 30 2018, 12:16 PM

There's no suggestion for glue stick or drying in the XSTRAND profiles
Also no suggestion for drying in PA-CF

Hob removed a subscriber: Hob.Oct 30 2018, 12:45 PM

I have a couple new tests to report. Two from Friday of last week working with ABS, I printed a pretty dang nice Rocktomech that was scaled up 250% I am pretty sure. Only real imperfections on the Rocktomech were the Z-Seams being a bit prominent but considering nozzle diameter I would say they are probably about as small as we can get them. I am honestly just trying to push the toolhead with as much difficult stuff as I can since I know that is what customers will be doing so I know exactly what it's limits are. I also tried to print two animal chess pieces that I maxed out on the build platform but flew to close to the sun on that one because even with the enclosure one of them warped up off the bed and ruined the print about halfway through. If I had done one at a time or added additional bed adhesion it probably would have come out great since the bottom part looks really nice aside from slight warping. Both of these prints were done using the standard detail profile with the only change being the fans were turned off.

Then yesterday I maxed out a cool alien bust model that was again picked to see how far I could push the toolhead with small details. Print was done in PETG on high speed profile and the model was scaled up by 250%. There was some Z-Seaming on the model but as with above I don't think it is going to be improved much past where it currently is. Overall I was very impressed with the way this model came out and can't think of any setting changes I would suggest.

Thank you @Galadriel! I am happy to see that. If you are looking for some extremes to test you could try something like baskets or crates that have a lot of open spaces along the wall, it forces a bunch of retractions and z hops, good for torture testing.
Happy Printing 😄

Updates pushed, fresh build should have updates incorporated.

adam added a comment.Oct 31 2018, 10:23 AM

Machine Type: TAZ 6 with adapter
Toolhead Type: Banded (HS)
Firmware Version:
Cura Version: 3.2.30 (rebuilt this morning)
Material Being Used: colorFabb PA-CF
Estimated age of filament: from Angie's desk - probably 3 weeks~ish
Profile being used: High Speed (0.5mm)
STL and/or Gcode: https://www.thingiverse.com/thing:291817
Environmental Description: Open air

Getting some pretty bad stringing on the parts. I think the flow rate might be off since it looks like the infill is pushing the nozzle up and getting too much material built up. I will edit with pictures once the print is finished.

Note to everyone testing, we are now on Cura 3.2.31 please preform a fresh master build by following step 6

adam added a comment.Nov 1 2018, 11:32 AM

Wanted to let everyone know about a time line update: we have two weeks to finalize profiles so we need your feedback by then. Please, please remember to rebuild your Cura and clear your cache to make sure you are testing the most recent/accurate profiles.

Noticing PA-CF profile moves bed forward at 110c

Machine Type: LulzBot Mini 2
Toolhead Type: Banded Tiger
Firmware Version: 1.1.9.17
Cura Version: 3.2.30
Material Being Used: GF30-PA6
Estimated age of filament: About 2 weeks since opening
Profile being used (please include any profile edits you made, no matter how small) : Standard 0.5 mm layer height. On some prints infill was changed to triangular with 70% infill for possible future strength testing.
STL and/or Gcode (add as file, web link, or shared-j location):

/home/aleph/shared-j/Documents/Creative/3D_Models/Made_By_LulzBot_Designs/General_Prints/Wrench/Wrench19mmx25mm_Closed_Small.stl
/home/aleph/shared-j/Documents/Creative/3D_Models/Made_By_LulzBot_Designs/General_Prints/Hook/Hook_WeightTest_02.stl

Environmental Description (enclosed or open air?) Open air printing
Pictures (to avoid data issues please remember to compress photos before posting): none available currently but will be edited to provide the pictures tomorrow.

General Feedback:

For how expensive the GF30-PA6 is, getting a dryer is a must. Unless constant printing was happening, you wouldn't be able to use up a full spool before it gets oversaturated. There were noticeable quality differences printing identical prints from one day to the next once the filament started steaming when extruding.

In general printing overhangs has resulted in a very poor surface finish, even when the filament was relatively new. I'm not sure if this has an effect on the part strength, but it's definitely not ideal aesthetically speaking.

@adam Should we still be on 3.2.31 or move to 3.2.32?

adam added a comment.EditedNov 1 2018, 12:16 PM

Machine Type: LulzBot TAZ 6
Toolhead Type: Banded Tiger with Adapter
Firmware Version: 1.1.9.26
Cura Version: 3.2.32
Material Being Used: PA-CF
Estimated age of filament: Several weeks
Profile being used : High-Speed 0.5 mm layer height. Retract amount upped to 5mm, retract speed raised to 25
STL and/or Gcode: https://www.thingiverse.com/thing:291817
Environmental Description: Open air printing
General Feedback:
Noticing some persistent stringing with this material. Lowered temp to 255 to eliminate, but the parts were weakened way too much at that temp. Will try hitting with a heat gun to see if that removes stringing the same way it does for PLA. If that is a legitimate recommendation I don't think it's too bad at these settings. I believe it's definitely worth looking into annealing this material if it increases tensile strength in Z direction.




Pictures are of parts printed at standard settings - adjusted settings exhibit less stringing

Machine Type: Mini with adapter
Toolhead Type: Banded (HS)
Firmware Version:
Cura Version: 3.2.32 (rebuilt this morning)
Material Being Used: IC3D PETG
Estimated age of filament: Roll has sticker saying 08/30/18
Profile being used: High Speed (0.6mm)
STL and/or Gcode: https://www.thingiverse.com/thing:655379
Environmental Description: Open air

Printed a basket with a lot of open spaces as a torture test. I flew a little too close to the sun on this one, connection points were thinner when they actually printed than I expected based on photos on Thingiverse. The stringing was only slightly worse than I was anticipating. The blobbing is entirely due to the model being too thin for the toolhead.

Mini 2 Banded 0.8mm PA-CF standard profile edit:

@anolen please give it a try, and let me know what you think. (Same goes for anyone else that would like to play)

adam added a comment.Nov 5 2018, 11:26 AM

Machine Type: TAZ 6
Toolhead Type: HS
Firmware Version 1.1.9.26
Cura Version: 3.2.32
Material Being Used: NinjaFlex
Estimated age of filament: old
Profile being used: Standard (did a couple prints turning up flow to 105% which looked better)
STL and/or Gcode: https://www.thingiverse.com/thing:2448971
Environmental Description: open
Results: Looks good, minimal stringing, but the top and bottom don't fully connect leaving line gaps. The perimeter walls also seem to delaminate due to small gaps between inner perimeter and outer. Turning up the flow to 105% seems to have closed them and made the walls solid, but I'm not sure if it's degraded the dimensional accuracy.

Hello Everyone,

We just pushed a slew of profile and FW updates. Please do a fresh build and be sure to update FW once connected.

Galadriel added a comment.EditedNov 9 2018, 8:45 AM

Machine Type: Mini with adapter
Toolhead Type: Banded (HS)
Firmware Version:
Cura Version: 3.2.32
Material Being Used: IC3D PETG
Estimated age of filament: Roll has sticker saying 08/30/18
Profile being used: High Speed (0.6mm)
Environmental Description: Open air

Tried out a Benchy and Antler with previous build of profiles and FW, both had some surface quality issues but I think a lot of that is the filament was waterlogged. When I extruded the filament outside of a print I noticed some rough texturing so I stopped printing with this roll of filament. I am still including the tests because I figure it is better to have the extra data.
For the antler I had to add a brim as well as supports (touching build plate), I have found that supports are particularly difficult to break off when using the banded. I know a lot of this is because it has better layer adhesion in general but I thought it might be worth suggesting an adjustment since a lot of my prints with supports ended up with ugly bottom surfaces. Maybe enabling the support interface roof or adjusting the support stair maximum step width would help with this?

3DBenchy

Antler

Machine Type: Mini with adapter
Toolhead Type: Banded (HS)
Firmware Version:
Cura Version: 3.2.32 (Firmware 1.1.9.28)
Material Being Used: ABS
Estimated age of filament: A few weeks
Profile being used: High Speed (0.6mm)
Environmental Description: Open air
Adjustments to profile: I added supports only touching build plate, and added a support roof to try and make removal easier.

I was quite pleased with the outcome of this print, there was a little bit of layer separation on the smaller areas of the model but considering it was printed in open air I would definitely say the layer separation was reasonable. The addition of the support floor made a HUGE difference in making the supports break away easily. It also improved the surface quality of areas where the model had supports in comparison to models printed without the support roof.




Just a heads up everyone, we plan to set a release candidate tomorrow. Get your feedback in now!

adam closed this task as Resolved.Nov 14 2018, 7:34 AM

Final profile tweaks and feedback have been moved into cura and it is capped - closing this as resolved.
If there are print issues, please open another ticket for specific issues.