[SOLVED] Retraction not Working

Ok, so I've been trying to diagnose this issue for the last few days, and I think I've found what the issue is, I only finished building my first printer 4 days ago, and I'm still making improvements so I'm very new to 3D printing and my assumptions could be completely off.

I've got my printer (Mendelmax based, 3.5mm J-Head MK-V Hot End, Gregs Extruder, Sanguinololu Board) printing relatively well, it's not perfectly calibrated yet, and I'm still learning a lot, but it's getting there. except for one major issue, Retraction causes a major amount of blobbing, and off course with multi start stop prints, no extraction leads to a lot of stringing.

I've tested retract and extrude using pronterface and can confirm that my setup successfully retracts and extrudes when applied manually.

I use Slic3r 0.9.8 and Marlin Firmware on the Sanguinololu board,

I'm using 3.0 mm ABS and my filament configuration is set to 2.95mm, with a multiplier of 0.85, and a temp of 220c
And my retraction is configured for 30mm/s and 1mm retraction and extra length 0mm.

By Watching the extruder I've made the following observation, the extruder does not visibly retract during printing, however it clearly performs the compensate retraction step and as a result extrudes a blob of melted plastic.

from reading up on G-Code I expect to see in the Slic3r generated G-Code something like this

G1 F1800.000 E-1.00000 ; retract
G92 E0 ; reset extrusion distance
---Do Some Movement here---
G1 F1800.000 E1.00000 ; compensate retraction


However what I do see is this.

G1 F1800.000 E1.71317 ; retract
G92 E0 ; reset extrusion distance
G1 X118.400 Y109.425 F7800.000 ; move to first perimeter point
G1 F1800.000 E1.00000 ; compensate retraction


Effectively what I'm seeing is that instead of retracting 1mm, the retraction step actually extrudes 1.71317mm, followed by a reset, and then a compensation extruding a further 1mm, in effect I would expect to see the Retract and compensate steps cancel each other out, i.e. retract 1mm, extrude 1mm with an effective result of 0mm actually extruded. instead what I believe I'm seeing is 1.71317mm + 1mm = 2.71317mm being extruded, causing a huge blob on each retract compensate step.

So am I simply reading the G-Code wrong or is this an actual bug in Slic3r? or is my configuration wrong.

Edited 2 time(s). Last edit at 01/24/2013 01:35PM by belatu.

What is the extrusion amount immediately before 1.71317? Is the 2.71317? If so then this is the proper code for 1mm of retraction.

The code you write with the -1 would only work with relative distances specified; most people let this default to absolute distances.

Also, the g92 is not really necessary for retraction. The code could have not done the reset, and just continued on with values 2.713317 and above.

The code before the retraction is.

G1 X119.650 Y119.650 E2.71317 ; perimeter
G1 X119.109 Y119.338 F7800.000 ; move inwards before travel
G1 F1800.000 E1.71317 ; retract

Ok, the G-Code makes more sense now since it's basically telling what position to retract to and not how much to retract. Time to look at the Marlin code to see what's going on there. or maybe just re-calibrate my extruder and use a slower speed for retraction...

I solved my issue with retraction, the first thing I did was re-calibrate my extruder. and updated the Marlin configuration with the new calibration data.

My Steps per unit calibration went from

#define DEFAULT_AXIS_STEPS_PER_UNIT {80, 80, 2560,900}

To

#define DEFAULT_AXIS_STEPS_PER_UNIT {80, 80, 2560,813}

The Second Step was to change the retraction acceleration

I Changed it from

#define DEFAULT_RETRACT_ACCELERATION 3000

TO

#define DEFAULT_RETRACT_ACCELERATION 1200

This seemed to make a big difference, but it was still not 100%

So back to slic3r.

First thing I did was reduce the retraction from 1mm to 0.5mm, that helped some but it was still not perfect. Then I reduced retraction speed all the way down to 10mm/s and enabled Lift Z setting it to 0.1mm

I'm currently doing a test build and Im seeing absolutely zero blobbing, and no stringing at all.

I'm printing 3mm Black ABS, at 255 C first layer and 220 C for the rest, with the bed at 95 degrees.

Thanks for pointing me in the right direction.

Actually intresting approach to retracrt, slow and little,
usually instructions are retract LOT if stringing

I've now printed 3 Buddhas, a new x Carriage, a bunch of small prints on one plate and a number of small prints with lots of moves etc and with the short and slow retractions I've seen absolutely zero blobbing, and no stringing even on prints that most people complain that they can not print without stringing, and I've printed anywhere from 50mm/s to 130mm/s.

My calibration is not yet perfect, but it's getting there, and after a few days of experimenting I'm convinced that small slow retractions work better than fast big retractions, especially when the biggest issue s blobbing.

I have to give a tip of my hat to belatu.. I've been having fits trying to print the lattice calibration pyramid. It's been a hot mess of blobs and strings and I was getting very frustrated. I used his recommended settings and it was almost perfect! I actually don't know if it could get any better. Thank you!!!!!
(picture attached)

Edited 1 time(s). Last edit at 02/12/2013 09:26PM by Knimrod.

OMG Thank you so much!@^&!*@^ I have been fighting my Prusa for weeks and changing everything I could and go figure this 10 minutes of tuning really did the trick!! Its amazing, I am also using Slic3r 9.8 with Pronterface (newest vers.) and mostly print off thingiverse but again, thank you!!! -Pb