• Welcome to the new Distech Automation Forum!
  • Feel free to post questions, comments, and feedback.
  • Ask us anything!
Hello There, Guest! Login Register


Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Dual color prints and tuning the settings
#1
All right, as I've had some minor problems with the retraction and Cura-settings, I thought that I'd post the results I'm getting at different stages of tuning.

So here's the first DC print: https://www.dropbox.com/s/e6hygljygijf7k...6.jpg?dl=0
PLA @205 C, almost untouched Cura-settings that came with the purge trail model.
There's some underextrusion and problems with retraction on the side, unseen in this picture. But not bad first DC print  Smile.

After tweaking the retraction and stepper driver vref (http://distechforum.com/showthread.php?tid=118), I printed Benchy: 
https://www.dropbox.com/s/xx43fl5g337gd3...6.jpg?dl=0
https://www.dropbox.com/s/94611id1uzzr2u...2.jpg?dl=0
https://www.dropbox.com/s/ajh2ge81ed4t6d...0.jpg?dl=0
https://www.dropbox.com/s/tu7kr7mdrggxh9...8.jpg?dl=0
https://www.dropbox.com/s/mkqplrawcnmhky...9.jpg?dl=0

ABS @240 C. Retraction length 2mm, general printing speed lowered to 40mm/s. Dual Color aspects of the print are very good, only a couple of blue dots in white areas, probably bits that got stuck to the nozzle and the detached during printing. It seems that the blue leaves a bit of the color on the nozzle and gives the white a bit of blue, so it's not perfectly clear. Would need to bump the purge tower size quite a lot to get it clean.

Otherwise, the print has some uneven surfaces and one spot where it seems that retraction might have failed. Not sure if this is because of Cura or my printer (rebuilt Cube2, i.e. FrankenCube); I designed new X-carriage and I have only used Prometheus with it so I can't tell if it's the cuplrit. It's quite solid, but still it's a printed part... anyway, more tweaking to the settings should shed some light into the matter.

Lastly, a Simplify3D test print:
https://www.dropbox.com/s/o3da88dlze2bqj...1.jpg?dl=0
https://www.dropbox.com/s/8az6h4chvz389r...9.jpg?dl=0

As I havent used Repetier/Cura before, I thought about other slicers I usually use:
- Slic3r supports multiple extruders, but no purge tower available. Would need to manually do that.
- KISSlicer also supports multiple extruders, but to actually map the objects to different extruders, I would need to buy the license. Still thinking about that, but for the time being cannot be used to do dc prints
- Simplify3D. I was curious about the problems with Prometheus and decided to see how bad the results are.Turns out that with bit of g-code tweaking, not bad. On the print above I didn't have all the settings properly dialed down, for some reason the purge tower was skipped during the first change to white, which is seen on the second picture (blue bleeding to the white border on the first layer). But after that, it did the material changes as it should. Also there's some underextrusion ongoing with the white, but that's probably because the Dual Extrusion wizard did some unwanted stuff at the first tries and I went and tried to do stuff manually -> the second profile was Default, with 0.9 extrusion multiplier.

Need to test Simplify3D more, but generally I'm more familiar with it so hopefully I'll get it tuned properly.
 
Reply
#2
Ok been testing different slicers and battling some problems.

As I got my a part cooling fan designed and printed (used Taulman Nylon 645), I tried some PLA.
https://www.dropbox.com/s/t4vhjiuftt6n71...9.jpg?dl=0
https://www.dropbox.com/s/tyqjsevr3q9nx2...2.jpg?dl=0

Here are 3 results, from left to right sliced with Cura, Simplify3D and KISSlicer 1.5 (bought the license):
https://www.dropbox.com/s/pesem20txpuils...5.jpg?dl=0
https://www.dropbox.com/s/1daie0tquwrk90...8.jpg?dl=0

Settings: 2mm retraction @80mm/s, 205 C temp (210 C on KISS version), cooling on.

The overall cleanest version is the Cura's as it keeps retractions to a minimum. Retraction is a problem with all of the prints, both that it's too long (horrible jamming as is seen especially on the yellow parts) and too short (stringing).
Looking at the S3D part's legs and back side it's clear that its not feeding material in a consistent way. Actually that print was the 4th S3D print that actually finished, first print was set to to use 2.6mm retraction and after few minutes I got a jam so bad that I had to stop the print. Luckily I got the jam freed without needing to open the whole thing. Lowered retraction to 1.8mm to get this result.
Black PLA seems to fare better. The yellow material came with Prusa MK2 and has been working ok on Prusa.

KISS and S3D are configured to do retraction very aggressively and those settings has been working on other printers, even with PLA, quite well. I'll have to investigate how to mitigate this problem if I want to print PLA. Cura's approach of minimal retractions (only when crossing borders) seems to help, but even then there's few places where there's clearly material missing. Maybe lowering temperature to 190 and/or trying to dry the filament. I've dried Nylon in oven and it works, but the problem is that the oven in my apartment doesn't have very precise temperature control and PLA might get too hot... maybe sauna Smile?

Other problems: only Cura does material switches in a way that the purging works. The problem with S3D and KISS is that if only one material is needed, no material switch is done -> gaps in purge trail. This causes results like this:
https://www.dropbox.com/s/cogv5w3e99mo3f...8.jpg?dl=0
https://www.dropbox.com/s/b9xvwd63rzzxpk...9.jpg?dl=0

So far the tangled mess hasn't caused real issues during printing, but it's far from optimal. I've been thinking about using the layer switch gcode to maintain some kind of scaffolding for the purge trail to adhere to.
Also KISS has quite limited options for purge tower, the cylinder on the pic above is it basically.

I'll switch to PETG to see how well it works.
 
Reply
#3
Ok getting problems solved (new ones emerging Smile ): 

1. Retraction jams. Added some silver thermal paste in the hotend heatsink and nozzle threads. Seems to work much better, currently @2.5mm retraction and even the stuck-happy yellow seems to work fine. Apparently the thermal paste improves the cooling enough to keep PLA happy.

2. Simplify3D and missing purge trail layers. Go it working. I use Layer change g-code to print a 'cage' around the purge trail area and this helps to keep the trail in it's place. Also added a raft on the start g-code so no need to add the purge trail model:
https://www.dropbox.com/s/128v65vv33hc1b...3.jpg?dl=0
https://www.dropbox.com/s/q8lkrj208gvnnx...7.jpg?dl=0
https://www.dropbox.com/s/95c7xr09rqx52r...9.jpg?dl=0

There's still problems; S3D seems to skip purge tower in some cases, there's one such layer visible here:
https://www.dropbox.com/s/fdvreg8sf3j0nw...1.jpg?dl=0
I might just turn S3D purge tower off altogether and modify the purge trail g-code to do another purge trail after filament change. This would also work with other slicers that don't have good options for purge configuration.

Also the obvious stringing and over-extrusion. I'm trying to hunt down the piece of g-code that causes some extra material to be extruded. It seems that it happens only on layers where there's tool head switch. If this is just a thing that S3D does, bummer.
https://www.dropbox.com/s/00yjzwnqzi2jyb...6.jpg?dl=0
https://www.dropbox.com/s/y3fxijdx22xse5...4.jpg?dl=0

I also had major problems with bowden tube getting sucked into the extruder and the bowden clip slipping. 
https://www.dropbox.com/s/nm776yf96j2nt2...5.jpg?dl=0
https://www.dropbox.com/s/fqlkg0yfiwehaw...6.jpg?dl=0
Apparently the constant jams earlier had caused the tube to start getting loose and after a while sliding in and out. The tape I had on the tube, as per instructions, didn't help much, those steppers do have torgue.
Other extruder is fine. The problematic one is the one with the yellow PLA.

To overcome this problem, I designed my own tube clamps that replace one part of the extruder assembly: 
https://www.dropbox.com/s/66zy2xqslye6bc...0.jpg?dl=0
https://www.dropbox.com/s/v52k800skli9gd...6.jpg?dl=0
https://www.dropbox.com/s/h39zuqgzma6jh1...0.jpg?dl=0
https://www.dropbox.com/s/50j5c5fato14bl...4.jpg?dl=0
 
Reply
#4
I switched off S3D's purge tower and modified the tool change g-code print more purge tower. Works quite well.
https://www.dropbox.com/s/hoq6pw9ub773xm...5.jpg?dl=0

No more odd skipping of purge tower layers:
https://www.dropbox.com/s/l2cx7as25tmhc2...5.jpg?dl=0
https://www.dropbox.com/s/4debcwe2sc3lr3...6.jpg?dl=0

Still, there's some oddities about layers, it seems like there's some over extrusion ongoing still... I wonder if it's because of my layer change g-code or something else. It shouldn't be S3D getting confused about filament position as I'm using relative extrusion lengths. By the way, S3D runs layer change g-code twice for each layer, because each layer is processed for each extruders. This makes it bit difficult to feed proper amount of filament. I'll try with KISSlicer next, hopefully it'll work better.

Anyways, I wanted to try something else and switched to PETG. Julia vase with red and clear PETG (21h print):
https://www.dropbox.com/s/udme13ix1kw070...3.jpg?dl=0
https://www.dropbox.com/s/udme13ix1kw070...3.jpg?dl=0

So PETG works fine regards the filament switches. The purge tower and cage need some tuning: as seen in the picture, there's these bits attached to the print and on the build plate and they come from the purge tower cage. These are cause by the cage g-code printing too fast and material not sticking properly -> random bits that get pulled to the print occasionally. Need to tune the g-code more.

Also for the vase, the cage was just wasting filament as there's two colors on each layer, hence the purge tower is built all the way.
 
Reply
#5
Can you share the Simplify3D settings you are currently using?  I haven't used Repetier/Cura for a long time. I also had jam problems with bowden tube getting sucked into the extruder and the bowden clip slipping. Apparently the extruder 1 did not retract on the first print; so when extruder 2 kicked in it was an instant jam in the Y-section. So its pull everything apart again.
 
Reply
#6
Sure, here's the FFF profile I'm using: https://github.com/spegelius/3DModels/bl...us_PLA.fff
Of course you need to change settings like bed size etc. to match your printer after importing it.

Yeah, I've had that Y-section jam once and it wasn't nice. Probably the main cause of later slipping. I usually check the resulting g-code to make sure which extruder is going to be loaded at the beginning. Latest KISS g-code actually unloads the filament because KISS always does extruder selection first so it doesn't matter which filament is loaded.

I thought about drilling small holes on the Y-section so I could check which filament is currently loaded.

Edit: here are the WP g-codes: https://github.com/spegelius/3DModels/tr...de_scripts
 
Reply
#7
Been trying to tune KISSlicer and PETG. Relatively good results so far, been printing @235 C. However, it seems that small details require more temperature:

https://www.dropbox.com/s/75ym7nttxc2fx7...2.jpg?dl=0
https://www.dropbox.com/s/0u5c7lj7fbql0b...3.jpg?dl=0
https://www.dropbox.com/s/clmg554qmz3ujb...3.jpg?dl=0

I'm going to bump temperature to 245 C to see if it improves. Also for this print there was no KISSlicer generated prime tower, only what I print in the Select new extruder-gcode.
I've noticed that getting that tuned properly is quite hard and there's one thing that can be a bit of hit and miss: retraction. It might happen in the g-code before layer change and extruder change, but it seems that in some occasions it doesn't. If it happens, all is fine as the custom gcode handles them properly (prime at the beginning and retract at the end). However, if there's not retraction before the custom gcode (and of course prime afterwards), this causes extra material at the beginning of custom gcode and gaps afterwards.
This actually is a problem with getting Prusa Slic3r to work with this custom g-code, the retractions and z-height changes don't match.

I've done some gcode file post processing in the past, I hacked S3D and Slicer to work with BFB style printers (code here: https://github.com/spegelius/CubePostprocessor). It shouldn't be too hard to add this custom gcode in proper places and see that the retractions are done in proper places. I might try that next. Also it would be easy to make the purge trail position easily configurable.
 
Reply
#8
Latest tuning to KISSlicer settigs seems to be good:

https://www.dropbox.com/s/bt4v5ry0njxpbu...8.jpg?dl=0
https://www.dropbox.com/s/07r3hwsipchv9h...5.jpg?dl=0
https://www.dropbox.com/s/815pyp423m1bvl...2.jpg?dl=0

Clear PETG with red isn't clearly the best option for dual color prints, but anyhow, in this case retractions are working properly. Current settings have both prime tower and prime trail in use, which seems to stabilize the extrusion amount.
 
Reply
#9
Smile 
(04-25-2017, 06:40 PM)spegelius Wrote: Ok getting problems solved (new ones emerging Smile ): 

1. Retraction jams. Added some silver thermal paste in the hotend heatsink and nozzle threads. Seems to work much better, currently @2.5mm retraction and even the stuck-happy yellow seems to work fine. Apparently the thermal paste improves the cooling enough to keep PLA happy.

2. Simplify3D and missing purge trail layers. Go it working. I use Layer change g-code to print a 'cage' around the purge trail area and this helps to keep the trail in it's place. Also added a raft on the start g-code so no need to add the purge trail model:
https://www.dropbox.com/s/128v65vv33hc1b...3.jpg?dl=0
https://www.dropbox.com/s/q8lkrj208gvnnx...7.jpg?dl=0
https://www.dropbox.com/s/95c7xr09rqx52r...9.jpg?dl=0

There's still problems; S3D seems to skip purge tower in some cases, there's one such layer visible here:
https://www.dropbox.com/s/fdvreg8sf3j0nw...1.jpg?dl=0
I might just turn S3D purge tower off altogether and modify the purge trail g-code to do another purge trail after filament change. This would also work with other slicers that don't have good options for purge configuration.

Also the obvious stringing and over-extrusion. I'm trying to hunt down the piece of g-code that causes some extra material to be extruded. It seems that it happens only on layers where there's tool head switch. If this is just a thing that S3D does, bummer.
https://www.dropbox.com/s/00yjzwnqzi2jyb...6.jpg?dl=0
https://www.dropbox.com/s/y3fxijdx22xse5...4.jpg?dl=0

I also had major problems with bowden tube getting sucked into the extruder and the bowden clip slipping. 
https://www.dropbox.com/s/nm776yf96j2nt2...5.jpg?dl=0
https://www.dropbox.com/s/fqlkg0yfiwehaw...6.jpg?dl=0
Apparently the constant jams earlier had caused the tube to start getting loose and after a while sliding in and out. The tape I had on the tube, as per instructions, didn't help much, those steppers do have torgue.
Other extruder is fine. The problematic one is the one with the yellow PLA.

To overcome this problem, I designed my own tube clamps that replace one part of the extruder assembly: 
https://www.dropbox.com/s/66zy2xqslye6bc...0.jpg?dl=0
https://www.dropbox.com/s/v52k800skli9gd...6.jpg?dl=0
https://www.dropbox.com/s/h39zuqgzma6jh1...0.jpg?dl=0
https://www.dropbox.com/s/50j5c5fato14bl...4.jpg?dl=0

Hi,
I've been having the same slipping issue with one of my extruders, could you post the STL for the tube clamp please?
Edward
 
Reply
#10
(05-11-2017, 03:44 AM)CobaltGriffon Wrote: Hi,
I've been having the same slipping issue with one of my extruders, could you post the STL for the tube clamp please?
Edward

Sure thing, I uploaded it to thingiverse. http://www.thingiverse.com/thing:2312435.
 
Reply
  


Forum Jump:


Browsing: 1 Guest(s)