Extruder Acceleration - Holy Crap!

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

Extruder Acceleration - Holy Crap!

Post by zemlin » Fri Dec 23, 2016 6:15 pm

Just wondering who has messed with acceleration settings for the Extruder vs XY movement.
I tried a little test part today in an effort to figure out some settings for improved print quality. I set XY accelerations to 100. I got bad flaring in the part at any direction change - so i started messing with extruder acceleration.

I didn't get close to a flat part until extruder accel was set to 1 (XY still at 100). Extruder at 1 and XY at 200 looked almost perfect. I'm running ABS with a .5mm nozzle, fwiw. I also find it interesting that the low setting for extruder accel increased the build time significantly. With accels this low I'm getting very long times for a small part, but I'm just starting out and will be looking for upper limits where ringing starts to be an issue. With higher extruder accelerations the build time for my little test cube was 40 minutes. With the Extruder accel at 1 it took 66 minutes.

The improvement in print quality, however, is astounding. Based on what I'm learning with this exercise, I think I'll have to go back and reestablish all my baseline settings. At this point I'm dealing with unacceptable build times, but I'm sure there's a lot of room for improvement in that regard.

If you've messed with Extruder acceleration settings, what did you learn?

I'm going to continue to explore this and will share what I learn.

The Vertical and Horizontal blocks to the left were with high extruder accel values. The blocks to the right were both printed with extruder accel of 1.
All parts were printed with XY accels set to 100, jerk 10, speed 5200 (but it never gets there because the accelerations are so low).
IMG_2617.JPG

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 Dec 23, 2016 10:23 pm

i haven't really confirmed this yet, but based on what I'm seeing in my test prints and machine operation, I believe the XY motion is not allowed to over-run the extruder, but the reverse is not true - the extruder can over-run the XY motion.

It seems reducing your extruder acceleration and jerk settings will impact overall print speed and prevent over-extrusion at the perimeters of solid fills and narrow, which I have always seen in my prints. XY Jerk settings do noticeable impact motion when the accelerations are low.

Here's where I am right now
In S3D on the OTHER tab
Default print speed 5200
Outline underspeed 33%
Solid infill underspeed 66%
support underspeed 100%
XY Movement speed 18000
Z Axis Movement 1200

Here are the accel and jerk lines in my start script:
M201 X3000 Y3000 E25 Z900 ; Set acceleration
M205 X20 E6 ; Set jerk

At this point I am retracting 1mm, not coasting, and wiping 2mm. Haven't spent much time on these settings.
Printer operation is super smooth. Ringing is barely visible in the part.

I don't think I'm at the sweet spot yet, but it's been a very interesting exercise.

User avatar
ednisley
Posts: 1016
Joined: Fri Apr 11, 2014 5:34 pm
Location: Halfway up the Hudson
Contact:

Re: Extruder Acceleration - Holy Crap!

Post by ednisley » Sat Dec 24, 2016 2:40 am

zemlin wrote:the extruder can over-run the XY motion.
Marlin, at least, coordinates the position / velocity / acceleration values across all of the axes: they move with trapezoidal acceleration profiles limited by the least-responsive axis.

I think an extremely low extruder acceleration reduces the XY acceleration & velocity enough to reduce all the second-order effects (like viscosity & pressurization). You could get the same results by directly reducing the XY acceleration & velocity settings to unnaturally low values.

For example, my configuration has:
  • Max feedrate: XY = 450, E = 94 mm/s
  • Max accel: X = 5000, Y = 2500, E = 10000 mm/s²
Printing a 0.40 × 0.25 mm thread at (let's say) 50 mm/s requires feeding the 1.75 mm diameter filament at 2 mm/s into the extruder.

Accelerating from a standstill to 50 mm/s along the Y axis requires (50 mm/s)/(2500 mm/s²) = 20 ms.

Accelerating the extruder from 0 to 2 mm/s requires only 200 µs at 10000 mm/s², so the firmware reduces the extruder acceleration to 100 mm/s², which matches the Y axis time. If it didn't do that, the extruder would be laying down far too much plastic early on in the path and not enough near the end.

With the extruder acceleration at 1 mm/s², it needs (2 mm/s)/(1 mm/s²) = 2 s to get up to speed, so the Y axis slows down to match it.

The Marlin acceleration settings have interactions I don't profess to understand (and I've probably goofed some of the calculations), but the overall structure works pretty much like that.

BrookeDavis
Posts: 4
Joined: Mon Jan 02, 2017 7:33 am

Re: Extruder Acceleration - Holy Crap!

Post by BrookeDavis » Mon Jan 02, 2017 7:38 am

t seems reducing your extruder acceleration and jerk settings will impact overall print speed and prevent over-extrusion at the perimeters of solid fills and narrow, which I have always seen in my prints. XY Jerk settings do noticeable impact motion when the accelerations are low.

3dPrintingMD
Posts: 279
Joined: Fri Oct 02, 2015 5:37 am

Re: Extruder Acceleration - Holy Crap!

Post by 3dPrintingMD » Mon Jan 09, 2017 3:45 pm

I don't know how to say this without sounding incredibly lazy, but someone needs to keep testing this.

I'm always looking for better settings, and gave these settings a try and have produced some of the best prints that I can remember seeing.

Very nice work here Zemlin with the testing and experimenting.
M2 - V4, MIC-6 Build Plate, Astrosyn Damper's(X/Y), Rev. E, Geeetech LCD

S3D - FFF Settings https://forum.simplify3d.com/viewtopic.php?f=8&t=2367
Print Quality Troubleshooting https://www.simplify3d.com/support/prin ... eshooting/

Bratag
Posts: 439
Joined: Wed Jan 14, 2015 5:33 am

Re: Extruder Acceleration - Holy Crap!

Post by Bratag » Tue Jan 10, 2017 9:52 pm

3dPrintingMD wrote:I don't know how to say this without sounding incredibly lazy, but someone needs to keep testing this.

I'm always looking for better settings, and gave these settings a try and have produced some of the best prints that I can remember seeing.

Very nice work here Zemlin with the testing and experimenting.
Have to agree on this. Tried them myself and definitely see a difference, not only in print quality but I can notice it physically in how the printer moves. Travel seems much smoother

ksevcik
Posts: 56
Joined: Fri Oct 21, 2016 1:07 am
Location: Houston, TX

Re: Extruder Acceleration - Holy Crap!

Post by ksevcik » Wed Jan 11, 2017 1:54 am

What model is that? I'm working on installing the latest Marlin RC on my RevE to see how the new LIN_ADVANCE affects things, and comparing stock settings to your extruder accel to LIN_ADVANCE.

hybridprinter
Posts: 141
Joined: Sun Dec 21, 2014 8:04 pm

Re: Extruder Acceleration - Holy Crap!

Post by hybridprinter » Wed Feb 22, 2017 3:26 am

Whats the status of these settings?

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

Re: Extruder Acceleration - Holy Crap!

Post by zemlin » Wed Feb 22, 2017 3:38 am

hybridprinter wrote:Whats the status of these settings?
I've been too busy to mess with this much. I did get some very nice quality prints using slower extruder accels - less ringing primarily - but I think it really needs a systematic approach to compare very low XY accel settings to low extruder accel settings - and I haven't had time for anything like that.
For now I'm back to more standard print settings

This is what I currently run for ABS.
M201 X2000 Y2000 E1000 Z900 ; acceleration
M205 X20 Z1.0 E10 ; jerk

Obviously not using the extruder accel to limit anything. I didn't notice much of an impact until I get below 100.

Phil
Posts: 179
Joined: Mon Sep 19, 2016 7:49 pm

Re: Extruder Acceleration - Holy Crap!

Post by Phil » Wed Feb 22, 2017 2:05 pm

zemlin wrote:This is what I currently run for ABS.
M201 X2000 Y2000 E1000 Z900 ; acceleration
M205 X20 Z1.0 E10 ; jerk
Is the above 'jerk' actually the derivative of acceleration per time? (da/dt)
I just thought that would be a function of the components, nothing you could control.

Post Reply