Hole, inside and outside dimensions wrong

I've sent an email to the guys who look after Slic3r but I thought I'd better post here as well so I can post their response incase others are having the same problem as me.

The Problem. I have an ABSPrusa Mendel that I have had up and running for several weeks and it's print quality is excellent. I have one niggling problem left to address though.

When I print things, they come out the wrong size. This is not a scaling or axis calibration issue because the parts are all out by the same amount. Holes are 0.25 mm too small and outside dimensions are about 0.25mm too big. This is normally ok but is a real problem when parts ar specifically designed to give a tight fit, i.e captured nuts and bolts, holes for screws.

I think I've followed the correct calibration procedure.

1. Calibrate the extruder so that when told to extrude 30mm of stock, 30mm gets extruded. Measure the diameter of the stock.
2. Print the perimeter test piece: -Peg and hole test. Parts did not fit together Peg = 10.12mm, hole = 9.6mm
3. Print thin wall test: - Thin Wall Wall came out 0.55mm so set Filament multiplier to 0.76 (0.42 / 0.55)
4. Ran thin wall test and it came out at 0.42 mm which is correct for my 0.4 mm J Head Mk IVb for a 0.3mm layer
5. Ran perimeter test again. This time was better at peg = 10.14mm and hole = 9.77mm but still not right. Parts are also more brittle and top layer does not fill properly.

What am I doing wrong? 0.76 sounds like a big adjustment for the Filament Multiplier and I think it's had an adverse effect on my print quality. I don't know if my printer is producing extrusions wider than it should or if Slic3r is making some kind of mistake in calculating flow rates.

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9

It sounds like it is your multiplier and the reason the layer adhesion is weak is because the 0.3 layer height is to high for a 0.4 nozzle. Try a layer height of 0.25

Edited 1 time(s). Last edit at 06/28/2012 01:46PM by Sublime.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

So, are you suggesting that I keep everything as it is but go for a 0.25mm layer height i.e. the Filament multiplier should be 0.76? Or, go through my calibration procedure again but use a layer height of 0.25?

Sublime Wrote:
-------------------------------------------------------
> It sounds like it is your multiplier and the
> reason the layer adhesion is week is because the
> 0.3 layer height is to high for a 0.4 nozzle. Try
> a layer height of 0.25

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9

OK, time to move over the Skeinforge as I'm not getting anywhere with Slic3r as this does appear to be an issue that has already been discussed by several sages in Reprap with no successful conclusion.

I have an ABSPrusa Mendel with a 0.4 J Head Mk IVb head and I'm extruding 3mm PLA. The machine was bought in its entirety from thereprapkitstore.co.uk. If there is anyone out there using such a machine with Skeinforge, can you please let me have your config files so that I can state experimenting? Either zip and post them here or IM me.

Thanks in advance.

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9

Hmmm,

I've just printed a peg and hole combination and they do seem a little out! I think we will have to wait for an update to slic3r and see if this addresses the problem.

How you getting on with Skeinforge? I might have to try that also!

Keukpa

welcome to FDM printing.

this is a universal problem for all machines that do not have some type of software correction.

The important dimensions are your OUTER dimensions. The reason holes come out smaller than designed is that the extrusion is pulled toward the centre of a hole by the nozzle while it is printed.

ways around this:

design a test piece with the desired hole and modify the hole diameter until it fits what you want it to fit

use skeinforge stretch and tune it until you get proper results

or the best solution in my opinion, check out nophead's polyholes. read his blog and find out why this is happening and his simple and universal solution. Then, help write a script for skeinforge so that all holes are converted into polyholes and come out the proper size.

Thanks for your response. I fully understand that I have built a machine for very little money that runs using free software and that my expectations of accuracy are adjusted accordingly. I realise that fdm, as a process, has fundamental limitations in accuracy. What I am trying to point out in this post and the issue raised in GitHub for slic3r is that the Gcode files generated by 084 also throws away potential accuracy because it is making a mistake in calculating the external perimeter. The path it chooses does not take into account some or most of the filament width. All solutions that I've tried up to now just work on reducing the filament widt. While helping reduce the size of the problem, they will not get rid of it until slic3r is fixed.

aplavins Wrote:
-------------------------------------------------------
> welcome to FDM printing.
>
> this is a universal problem for all machines that
> do not have some type of software correction.
>
> The important dimensions are your OUTER
> dimensions. The reason holes come out smaller than
> designed is that the extrusion is pulled toward
> the centre of a hole by the nozzle while it is
> printed.
>
> ways around this:
>
> design a test piece with the desired hole and
> modify the hole diameter until it fits what you
> want it to fit
>
> use skeinforge stretch and tune it until you get
> proper results
>
> or the best solution in my opinion, check out
> nophead's polyholes. read his blog and find out
> why this is happening and his simple and universal
> solution. Then, help write a script for skeinforge
> so that all holes are converted into polyholes and
> come out the proper size.

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9

Looks like there was a bug and it has been fixed and will be available in the next release.

Edited 1 time(s). Last edit at 07/03/2012 03:23PM by Sublime.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I have just built my own Slic3r Dev0.8.5 from the master and verified that the issue is fixed. Well done and thankyou

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9

Can you share the fix here? I think I have the same problem.

Thank you

---

Ernesto Pareja

dcctrains.netne.net

You have two choices:-

1. Wait until 0.8.5 is released.
2. Follow the relevant instructions here to get and build the latest master.

I did 2.

I don't know when they will release 0.8.5

epareja Wrote:
-------------------------------------------------------
> Can you share the fix here? I think I have the
> same problem.
>
> Thank you

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9