Current Slic3r build won't do the "calibration cube" wall thickness - is it me ?

There is an old article that describes the method, and has a link to the cube .stl (I guess it's nothing new to people here)
[print.theporto.com]

I attempted to do this calibration to fine-tune an filament I am already printing petty well with.

The issue is, that latest build of Slic3r 1.3.1-dev does not seem to even try to make the walls 1mm, it's more like 0.4mm (nozzle width)
Is there something I miss ? - some new setting ? - or a bug during development ?

First, a 20 mm cube is too small to calibrate anything. The plastic previously deposited will still be hot when the nozzle comes around to deposit another layer of plastic. The result will be a sloppy looking print, and when you try to measure the wall thickness it will measure thicker than it should because the layers aren't stacking neatly on top of each other. Second, if you have a 0.4 mm nozzle, you can't print a 1mm wide line.

In slic3r you control the line width under the print settings>advanced menu. Be sure you're using slic3r 1.3- the latest developer version. 1.29 was very buggy and many of those bugs have been fixed in 1.3.

The nozzle diameter determines the line widths and layer thicknesses available to you. With a 0.4 mm nozzle, layer thickness should be under 0.3 mm or print will have weak interlayer bonding. 0.2 mm is a typical layer thickness for a 0.4 mm nozzle. Line widths should go from the nozzle diameter up to about 50% greater than the nozzle diameter, so 0.4- 0.6 mm.

Calibrating an extruder based on a single line width is prone to errors. Print a reasonably large object with 3 or 4 perimeters equal to the nozzle diameter. So, you might print a part that's 50-100 mm on a side, with line width of 0.4 mm and wall thickness of 1.6 mm. When you measure that print, you will divide the measurement by 4 which will give you the average line width which is likely to be closer to correct than trying to measure a single line width wall. Also, when you measure, measure at several points along different walls of the print and calculate the average value. Be sure to measure at least a few mm above the bed for most accurate results because heat from the bed can distort the first few layers of a print. Also, measure away from the corners- corners tend to come out thicker because of the interactions of acceleration, jerk, and extrusion.

Your calibration doesn't mean anything if you don't use the actual filament diameter when you slice the test part. Measure the filament diameter in 20-30 places and calculate the average value, then use that average value when you slice. When you slice and print leave the extrusion multiplier/flow control set to 100%. After you have calibrated, in the future you'll either need to print a test part with each spool of filament and then determine the flow/extrusion multiplier setting to use, or you'll measure the filament diameter in 20-30 places and calculate the average and use that when you slice or at print time if you use volumetric extrusion. Write the extrusion multiplier or average diameter on the spool and use those values when you slice/print from that spool. While you're doing that, weigh the new spool and mark the empty spool weight on the spool, too. That way you can check the remaining filament by weighing the spool before you start a long print.

Of course, this stuff is really only necessary if you are trying to print accurately, for making close fitting mechanical parts, etc. If you're printing Yoda heads you can just use the nominal filament diameter and be done with it.

Edited 1 time(s). Last edit at 07/31/2018 10:54AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

First of all, thank you for a very complete answer. I knew a lot of it from before (it's about 10 years ago I built my first Prusa) , but that's a still a very good answer

The reason for my question is:
I am printing mechanical parts, that need to fit well together. I am printing a batch of 20 of each part (3 parts)
Due to an failed print (OctoPi % printer lost power as my youngest son switched off the UPS) - so I discovered that some thin walls (two-line-width) in one of my prints that I always thought were "perfect", were not sticking together.

This print has four solid layers on bottom , and a footprint of ~16x5cm increasing extrusion multiplier was a time consuming approach, (those layers are slow) and I increased too much, and clogged once too.
So then I wanted to try this method to get the width work.
As you point out 1mm for a 0.4mm nozzle is not very likely to happen.
I increased the extrusion multiplier to 1.05
An hour ago another print got done, and it's perfect. I believe that 5% were about right.

Thank you.