Slicer-induced (Simplify3D) extruder jam

[Jake]

Hi everyone.

I've been calibrating Simplify3D and am having trouble with extruder jams.  The extruder jams at about the same place each print.  After it happens the extruder stalls, and you can clear the jam with a little pressure on the filament while running the load filament script.  I am currently using budget black ABS, which isn't the best but Makerware doesn't seem to have a problem.

I'm running a 3-in-1 extruder adjusted appropriately, so I don't think it is a filament diameter issue.  I'm wondering if it has something to do with the Wipe and Cost settings, since I don't think MW has an equivalent in the slicer, unless it does under the covers.  For reference, some relevant settings:

Retraction Distance 1.30mm (same as MW)

Retraction Speed 1500.0 mm/min (same as MW 25mm/sec)

Temp 230c (same as MW)

Wipe 5.00  (no MW equivalent that I know of) 

Coast 0.10 (no MW equivalent that I know of)

I've also tried both concentric and rectilinear fills.  With concentric fills, the smaller and smaller diameter circles get too fast. (known motion planner limitation)  I was thinking that this might be causing the jam, but it also happens with rectilinear fill in the same place.

Anyway, rather than trial and error, was wondering if the Simplify3D experts might have an idea about what's causing my issue.  Thanks for any recommendations.

 

Aside, does anyone know if its possible to tell Simplify 3D to slow down small loops and curves?  Would make using the concentric fill a smoother experience, given the limitations of the processor and the fact that the motion planner doesn't like small tight curves.

[Joseph Chiu]

In some cases, excessive retraction as you get near regions full of small infills can chew up filament.  Reducing the retract distances (or eliminating them altogether) could be what you need.

Another possibility that I've encountered is that rapid changes in flow rate sometimes has caused the thermal regulation of the hot-end to go into a strange state where the TC reads the correct temperature, but the filament is itself not at the correct temperature to be able to feed.  When the filament is fed at a steady stream, it flows smoothly, but when the feed is slowed down or even stopped, when you try to resume, the filament will not feed, or will offer very high resistance to feeding for a second or two before feeding smoothly again.    I encountered this reduced-flow behavior with my ToM, and had experienced a really bad case of it with a 2X.  My explanatory description may not be an entirely accurate description of what is happening, but the behavior is definitely as I described it.

Finally, with some slicers, the temperature is moved up and down throughout the print.  The PID is reset when a large adjustment in temperature occurs, and the resulting regulation error can cause the nozzle to be too cold for extrusion. It's probably not the case here, but wanted to mention it.

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To unsubscribe from this group and stop receiving emails from it, send an email to makerbot+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Jetguy]

My thought is that the extruder is marginal and Makerware babies it. I doubt that S3D is producing some code that jams the extruder.

 

I understand your frustration and I'm not trying to say you are wrong in your observation, just blaming it on the slicer doesn't solve the problem.

 

I know the STL but can you post your FFF file? I have S3D handy and can take a look.

[Jake]

Hey, no frustration.  If this helps identify another problem with the extruder than all the better.  

Some notes on the extruder-- its a 3-in-1 with the stiff spring installed.  With the original weak spring, I could stall the extruder by pinching the filament, and it even with too many turns on the adjustment for my liking. With the stiff spring, it only took maybe 1/2 a turn past "flush"

It is a QUBD nozzle, not the original Makerbot one, but it has been working fine otherwise.  Filament is as I said, budget black ABS.  Can't wait to burn through it so I get get to the good spools from ToyBuilder. :)

FFF is attached.  Lots of options, so maybe I missed something obvious.

[Jetguy]

Is this a 2 or 2X?

[Jake]

Rep 2X, Right Extruder

[Scottbee]

I had to re-do the extruder roller pressure on both of my 2X extruders.  Don't know if they "sagged" with time, vibrated (I don't believe they used nysert set screws or Loctite) or were set "light" from the factory.

[Jake]

Scott, I've got the 3-in-1 extruder upgrades.  I think that the problem with the stock extruders is that the lever mechanism pushes the spring past their critical distance, causing damage to the sprint with time. (Coil crush)

At least this is what happened to my extruders with time.  See Jetguy's post here: https://groups.google.com/d/msg/makerbot/SuzXohTGeZs/HCegADvOybkJ

The 3-in-1 doesn't suffer from this design issue.

[Scottbee]

Sorry.... I missed the post where you mentioned that you had the 3-in-1.  My bad.

[Scottbee]

(also as a side note... I don't use the release lever.... )

[Jake]

So it might be a retraction problem, possibly exacerbated by heat. Its possible that S3D uses more retracts, especially when doing concentric fill.  

I just at a Makerware print fail that had heavy use of retract/restart.   I was able to get it to complete by starting again and leaving the top open.

On Friday, 7 February 2014 10:43:31 UTC-5, Jetguy wrote:

[Philipp]

Jake,

I had a similar problem when I switched to Simplify3D, the nozzle was clogging half way during the print ( with ABS ), Stepper skipping steps in faster sections until total blockage and shitty flow in general.
In the end bed leveling was to blame ( even though it only started showing after 10-15min printing): Somehow S3D and RepG didn't agree with the endpoint, and the bed was too high for S3D. By the time I had it re-leveled the nozzle must have clogged a little bit with overheated ABS and hairspray, just enough to restrict the flow enough to cause trouble. After proper cleaning, and bed leveled for S3D it never happened again; I printed at least 10 hours after that ( same filament ). Never touched RepG or Makerware again btw.

Just something to think about

[Jake]

This is an interesting point.  I don't think its leveling, as I've got that dialed in pretty well and a glass build plate... however, you're point that early back pressure can manifest itself later is an interesting idea.

I am not a fan of the S3D priming sequence, which pretty much leaves a 3/8" blob of ABS and then drags the nozzle out of it.  Seems like this could start the print off badly... maybe that catches up later?

Also I noticed that Enginwiz's ABS profiles use much less retroaction then I was using.  The S3D default is 1mm, Makerware uses 1.3 default.  Enginwiz is using 0.25mm.  I intend to try Enginwiz's profiles soon for both ABS and my first attempt at printing Nylon.

[Scottbee]

Interesting.  The way S3D is set up on my printer, the initial extruding is done "off of the build plate" and it just puts and extruded string down into the printer pan.  Then the nozzle traverses across the build plate and starts the print or skirt.  I didn't think the initial extrude was supposed to be done on the actual plate.

[Fozzie]

I'd love that because, for me, it does the purge in the near left corner against the plate just as Jake describes.

[Scottbee]

On my 2X it appears to go max -Y and -X.  That puts it well off of the plate (for the left extruder).

[Scottbee]

Yeah... this is in the Starting Gcode:

G1 X-130 Y-75 Z30 F9000 ; move to wait position off table

I would guess that you could edit the Start.Gcode to get yours off of the table too, assuming you have the axis travel.

[Jake]

Well, I altered the starting gocde to do a Makerware-style leader.  Also lowered the retract distance to 0.25mm, turned of coasting and wipe.

I was able to print a standard spool holder with these settings, but Wingcommander's retraction test piece, no go. still fails in almost exactly the same place.

Back to the drawing board.

[Jimc]

in s3d, the initial prime is supposed to be OFF the edge of the bed then it should do a wipe sequence using the edge of the bed to scrap off the purged filament. if its priming with the nozzle on the bed something is not right. 

[Jake]

Might be the Bottleworks build plate, having slightly different dimensions, and keeping the right nozzle on the plate even when full left and forward.

Even with a custom priming sequence, I still get the extruder jam on Wingcommander's test piece, so I'm currently stumped whats wrong in S3D that doesn't happen in Makerware.

[Jimc]

just to cover the bases, you do have the platform size all set correctly and the table flipped in the right direction in the preferences tab?

[Jimc]

if so maybe you cna measure your bottleworks table and make sure it matches your settings. maybe increase the setting on that axis by 10mm as well

[Jake]

On my machine,  full X-Negative Y-Negative (Full front, full left) leave the left nozzle on bottleworks' build plate.  Not sure if that's normal, but I think it is.

It would be easy to prime the right extruder at the home position or at X-postive Y-negative.  These two positions put the nozzle off the platform. but thats really just polishing up the g-code scripts.   Figuring out why the S3D print causes an extruder jam, while the same part sliced similarly in Makerware prints fine.

[Jimc]

ive been using s3d for quite awhile now and am pretty familiar with it but i dont run a MB, i have an m2 so i am only tossing ideas here for you. for the clogging, i cant see that being slicer specific other than the differences in some of the settings from MW to creator. have you tried increasing the temp? bump the temp up 10deg just to see what happens. or for that matter also try lowering it 10 deg as well. i would think higher would be better though. also try decreasing the extrusion multiplier a bit. i have used all the ooze control settings but never the wipe function. do you need it to wipe? i would shut that off if not. in your original post you said you have coasting set to .1. is that correct or dod you mean 1.0. default is 1.0. coasting can cause issues on very tiny detailed parts because it basically stops pushing out filament "X" distance before the end of a line to reduce nozzle pressure. sort of like releasing the pressure on a caulking gun to keep the caulk from continuing to come out. on small detailed items like in your picture the extruder wont build up enough pressure so you end up with small sections of line missing and unprinted. if all the lines going down are like that then you end up with a situation where the part isnt printing and could be mistaken for a jam.

[eddyogi]

I have been having a similar problem with S3D.  The bottom part of the STL seems to print well, apart from all the zits I still get.  But when it starts printing the curved top section, it starts to not extrude enough plastic.  Even the infill is very weak and flimsy.  I've tried it with resolutions up to 100 microns but I still get the same failure starting at the curved section.  Check out the photo.  The print at the very back of the photo was made using RepG and (apart form the zipper) printed fairly well.  I've tried changing the retract from default to 3.  I am going to do a bit more experimenting, but if anyone has a proper fix, I'm sure everyone will appreciate it.

Cheers.

[Jake]

Its a small sample size, but I just had a successful print by disabling "force retraction between layers", and another failed print subsequently with it active.

I have to do some more testing, but I do think my problem is retraction based.

[Jake]

Yeah so "force retraction between layers" is not my problem.  I was able to get the issue to occur with this option both on and off.  Also, I tried a range of temperatures, with no effect.  

Also a little analysis shows that the code generated by Makerware actually does more retracting than the code generated by Simplify3D.  So I don't think its a problem with an over abundance of retractions.

Still trying to figure out why.  Tomorrow might try getting rid of the retractions altogether and see if at least the print can complete without an extruder jam and air printing.

[Jimc]

jake just for the hell of it why dont you email s3d support and see if they can give you an idea of the settings that might cause that. include you fff file when emailing them. i will get a situation like that once in a while and it just comes down to my hot end being half clogged to tight. i get it if i switch back and forth with different types of plastics. sometimes they dont like to mix well together and the tube will have a coating of the first plastic i used on the inside and it never gets pushed out. it just stays there and cooks for hours getting harder and harder. it doesnt completely clog anything but it makes it tight enough to cause strange problems like what your getting. why makerware doesnt seem to do it i have no idea.

[Jake]

Ok so I think I found the issue, and its not something I'd expect.

"Adjust printing speed for layers below"... seems to be defaulted to 15secs. Makerware uses a 5 second limit for a similar setting "minLayerDuration"

I'm not exactly sure how/why this causes an extruder jam.  If I had to guess is that as the layers get smaller, the slicer is slowing down too much and the extruder is running very slowly, and fact that the plastic in the nozzle isn't extruding very quickly gives it time to fowl and jam.  This is budget black ABS after all.

I also upped "Allow speed reduction down to" 30%, which I think is similar to Makerware's minSpeedMultiplier.  The combination of these two changes seems to fix my problem.

[Dan Newman]

On 13/02/2014, 10:10 AM, Jake wrote:
> Ok so I think I found the issue, and its not something I'd expect.
>
> "Adjust printing speed for layers below"... seems to be defaulted to
> 15secs. Makerware uses a 5 second limit for a similar setting
> "minLayerDuration"

And, keep in mind, that those min-times are truly minimum times:
the slicer doesn't know your bot's acceleration settings. The slicer
may be assuming infinite accel/decel. I.e., the slicer may be
assuming that the layer is printed at the target speed from
start to finish. With acceleration, it's printed slower than
that. Possibly as much as 20 - 30% slower. So, cooking time
may be even longer -- long enough for heat creep to start
becoming a possibility.

Dan

[Jimc]

yeah your right jake. that completely slipped my mind. i guess it never dawned on me because i dont get heat creep in my extruder where that is ever an issue. anyhow glad you found the problem. 

[Scott K]

I know old post but I wanted to chime in on the BottleWorks build plate and replicator priming.  It is true, you can only get the right nozzle off the build plate in the front right corner of the bot (and just barely (2mm maybe)).  So I just have my prime line run the front of the plate about .5mm in from the edge of the glass.  I like to see the line anyway so I know that I have the bed nice and level with good adhesion.

That being said, you can modify your build plate dimensions in S3D for a Left or Right ONLY print to get a lot more space than the stock build plate.

With my right nozzle ONLY, I have the following build dimensions (with ~1mm of cushion in each direction just to be safe).

This is on a FF creator Dual, but I believe the dimension are the same as a Rep1.  I cut down my Z-axis to 145 to avoid a potential bind at the axis extreme.

On Monday, February 10, 2014 7:53:18 PM UTC-6, Jake wrote:

Might be the Bottleworks build plate, having slightly different dimensions, and keeping the right nozzle on the plate even when full left and forward.