Extruder Acceleration - Holy Crap!

General discussion topics
User avatar
zemlin
Posts: 453
Joined: Thu Feb 04, 2016 9:32 pm
Location: Indianapolis, Indiana

Re: Extruder Acceleration - Holy Crap!

Post by zemlin » Sat May 20, 2017 12:10 pm

More recently my thought process has been moving away from slow accelerations.
Some other testing I did has led me to believe that ringing on prints at points of directional changes is not the result of dynamic instability. I am in the midst of a big print job right now which is preventing me from further testing out my hypothesis, but I believe ringing is actually the result of speed profile mismatches between the extruder and the gantry axis. My settings are moving in the opposite direction discussed early in this thread. My extruder acceleration is currently set at 10000 and extruder jerk is 100. It's a guessing game for me since i don't know the inner workings of the controls on these machines.

I still get a little ringing at the perimeter start point, but hard corners are quite clean and surface finish overall is looking pretty good, and it's not costing me print speed.

User avatar
jimc
Posts: 2888
Joined: Wed Apr 09, 2014 11:30 pm
Location: mullica, nj
Contact:

Re: Extruder Acceleration - Holy Crap!

Post by jimc » Mon May 22, 2017 3:31 am

extruder acceleration is just affecting the extruder motor, as in the filament retractions etc. this will have no effect on ringing or anything like that. x and y accelerations are what is going to affect the ringing. those are set default at 3000. most of us that change it are around 1500 or so. seems a happy middle ground to help the ringing without making seams too heavy.

User avatar
zemlin
Posts: 453
Joined: Thu Feb 04, 2016 9:32 pm
Location: Indianapolis, Indiana

Re: Extruder Acceleration - Holy Crap!

Post by zemlin » Tue May 23, 2017 12:04 am

jimc wrote:extruder acceleration is just affecting the extruder motor, as in the filament retractions etc. this will have no effect on ringing or anything like that.
My observations don't agree with your explanation.

User avatar
jimc
Posts: 2888
Joined: Wed Apr 09, 2014 11:30 pm
Location: mullica, nj
Contact:

Re: Extruder Acceleration - Holy Crap!

Post by jimc » Tue May 23, 2017 12:47 am

what would filament extrusion have anything to do with ghosting artifacts? the ghosting is from the vibrations traveling through the machine during a sudden movement change. this was the whole reason the cory xy gantry system was designed by that team at MIT.

User avatar
zemlin
Posts: 453
Joined: Thu Feb 04, 2016 9:32 pm
Location: Indianapolis, Indiana

Re: Extruder Acceleration - Holy Crap!

Post by zemlin » Tue May 23, 2017 11:46 pm

jimc wrote:the ghosting is from the vibrations traveling through the machine during a sudden movement change.
This is precisely the assumption that I am questioning based on observations I've made which, IMHO, do not align with this explanation.

1) One of the first things I did with my M2 was install dampers on the X&Y motors. I was concerned that reducing the rigidity of the gantry drives by adding the elastic element would result in increased ringing. I expected an increase in the magnitude or a decrease in the frequency of the ringing artifacts, but I saw neither in my before and after test prints. I could not see any difference resulting from the change to the mechanical properties of the drives.

2) A while back i was running some test prints in an effort to optimize acceleration settings. At one point I set the X-axis speed and accel very low while leaving the Y axis high. The expectation was that ringing on the Y axis of the print would be impacted by changes to the dynamics of the X-Axis.
XY.jpg
What I found, however, was that changing the dynamics of the X-Axis movements had no impact on artifacts on the Y-face, but rather smoothed out the printing in the X-direction.

3) If ringing is due to the dynamics of the X and Y axes, one would expect the plastic extrusion to be a wavy line with constant width where ringing is occurring. Close inspection of many prints has show that the extrusion width varies along a fairly stable path. To me, this is an indication that the linear movements and the movements of the extruder drive are stair-stepping their speeds at different points resulting in variation of the extrusion width.
WavesVsWobble.jpg
KZ6_6089-1600.jpg

User avatar
jhaupt
Posts: 64
Joined: Fri Aug 14, 2015 9:48 pm
Location: Mattituck, NY
Contact:

Re: Extruder Acceleration - Holy Crap!

Post by jhaupt » Fri Sep 22, 2017 12:11 am

This is some incredible work zemlin. Great, great work.

I'm an M2 user but lately I'm commissioning a scratch-built large-format machine, and have been banging my head against a wall with some heretofore unseen artifacts. After weeks of eliminating potential causes, one of the only things I left left to look into is the extruder acceleration and jerk, which I never considered important before.

One of the issues is the variation in extrusion width that you've captured so well. BTW, if I didn't know any better I would say you took that with a telecentric lens, like one would have on an optical inspection system (?).

Anyway, it's been a few months and I wonder if you have any updates to say. I'll be playing with new combinations of extruder vs. XY acceleration and jerk starting tonight.

~Justine

User avatar
zemlin
Posts: 453
Joined: Thu Feb 04, 2016 9:32 pm
Location: Indianapolis, Indiana

Re: Extruder Acceleration - Holy Crap!

Post by zemlin » Fri Sep 22, 2017 3:29 am

jhaupt wrote:Anyway, it's been a few months and I wonder if you have any updates to say. I'll be playing with new combinations of extruder vs. XY acceleration and jerk starting tonight.
Thank you. I am a part-time professional photographer and have some good gear. I used a 105mm macro lens for that shot, and may have added some extension tube to get closer. I don't recall the details.

As far as an update, I've been too busy printing to play with theoretical stuff. Sorry!

User avatar
jhaupt
Posts: 64
Joined: Fri Aug 14, 2015 9:48 pm
Location: Mattituck, NY
Contact:

Re: Extruder Acceleration - Holy Crap!

Post by jhaupt » Mon Sep 25, 2017 9:43 pm

Okay, thanks for the reply though.

M3Jim
Posts: 19
Joined: Sat Jun 13, 2015 2:16 pm
Location: FL

Re: Extruder Acceleration - Holy Crap!

Post by M3Jim » Fri Sep 29, 2017 6:41 pm

I have noticed the play in the X-axis slide. I briefly looked into a better option for it, but I don't remember the details. I think there is a tighter tolerance and/or a wider slide assembly available. Has anyone else looked into this, or even upgraded?

Have a great day,
Jim

Tinchus2009
Posts: 1
Joined: Mon Apr 09, 2018 12:05 pm

Re: Extruder Acceleration - Holy Crap!

Post by Tinchus2009 » Mon Apr 09, 2018 12:30 pm

Hi. I have been dealing with bulge corners and so I found this post. I can confirm pretty much ob the observations shown bt zemlin.
Every piece of information out there states that bulge corners is a problem related to overextrusion when the hotend need to stop at corners and I think everybody has told me to reduce speed and accel. I Did, gettim to values of 700 for X and Y, x/y jerk 10, extru jerk 0.3 with speed of 50mm/s And I can say bulge corners got even worst.
I started playing with pressure control, useing linear advance control settings. I got imprevments but at a cost of the extruder motor makeing horrible noises and my printer (I3) was really silent.
Las night I spent all night playing with linar advance values and settings till I found your post... I abandoned the pressure control aprouch and started again with accel settings and .... MAGIC!!!! I returned to a 1500 accel value for X and Y, and a jerk x/y of 15 but I reduced Extruder accel from the default of 5000 to only 100. The first tst print was... well, I guess perfect is not posible but, there were almost not at all any signs of bulge corners, just just very very little, no ringing. Also speed print went up :)
When Im talking that the bulge corners are gone... it is really difficult to detect them now, I have to look closely to every corner to see some very small defect there.
Then I tested adding the pressure lineal control... the bulge corners are gone 100%, not even with a magnification len I can see them... But the noise comeing from the extruder motor is so nasty, than I just prefer not to use pressure control at all, is not worthy.
My thoorical knowledge is not even close to your but I can say following the observation of my print test, that extruder accell has something to do here.
My test prints were done useing a section of a real world object, wich needed good "fitting" print details. That object in the past required big post processing work, for removeing corner bulge details that prevented this fitting into other object. With the "new" settings no post precessing is needed, the object just fits together at once perfectly. I have to say that other aspect of my printer are nicely calibrated (extrusion, steps, etc)
ould be interesting to see this post with deeper analisis and info

Post Reply