Deviation of 1mm  

Re: Deviation of 1mm

Andreas

The printer will need to be calibrated perfectly especially with regard to the filament feed; the filament will also need to be consistent in terms of diameter and hardness. Print accuracy will also depend on the model, the slicer and the slicer settings.

When required I can print parts that are accurate to about 0.1mm (or less).

For accuracy, it is very important to have the extrusion width set to the same size as the nozzle, print the outside perimeter first and print slowly. Unfortunately, this means that it is difficult to accurately print models which significantly angle outwards (overhangs), but model orientation is the key to most of those.

Peter

Re: Deviation of 1mm

By the default S3D settings provided by Josef, the Extrusion Width is set to 4.5 and Nozzle Width is set to 4.0. Should I set the Extrusion Width to 4.0? I also use the filament provided by Prusa. But I have not measured the thickness.

Re: Deviation of 1mm

Andreas

I note that Josef has all his default settings to print wider than the nozzle; except for when I am printing fast, I have them both the same setting.

Josef certainly has much more knowledge about this than me, but I note that Anderson (a RepRap "veteran") has recently made available his Cura settings in which he too uses 0.4mm line width...

Thing is, my understanding is that to get best accuracy, you have to reduce the unknown factors, which include material squishing out from the side of the nozzle.

I guess the only thing for you to do is to try out some test prints and use the settings which provide the best accuracy for you.

Peter

EDIT: Regarding the filament; it is reasonably safe to assume good quality filament has a reasonably consistent diameter, but when accuracy is required, you have to measure the diameter (in 2 directions 90 degrees apart) at many points over a good length (5 to 10 Metres) to ensure that all is OK and use the average of those values for the slicer filament diameter.

Re: Deviation of 1mm

So far, on my MK2 upgrade, the largest deviation I've seen has been roughly 0.25mm. This was sliced in Slic3r using Josef's defaults. My PLA is 9months old though and not stored in dry conditions so that may be skewing the results!

Re: Deviation of 1mm

Have I configured my Prusa i3 MK2 correctly, or is this a normal deviation?

I think your deviation is higher than expected.

I print using the default settings with Slic3r, and average quality filament (eSun mostly). I measured one of the parts I printed just now - the longest dimension of that part is 195 mm. The deviation on the printed part is part is less than 0.2 mm.

Re: Deviation of 1mm

For accuracy, it is very important to have the extrusion width set to the same size as the nozzle, print the outside perimeter first and print slowly. Unfortunately, this means that it is difficult to accurately print models which significantly angle outwards (overhangs), but model orientation is the key to most of those.

Peter

Sorry for digging out that old thread.

Peter can you please elaborate this? I mean the slicer will take the extrusion witdth into account? So if extrusion width is 0.45 it will move the head on each side 0.025mm narrow compared to a 0.4mm extrusion width.

Carsten

Re: Deviation of 1mm

Carsten

Theoretically, yes.

Bur consider the perimeters and the manner in which the filament extrudes. If you print from the inside perimeter to the outside, you have laid down a limiting ring; and minor deviation in extrusion width and "squish" can only go one way - outwards. The second perimeter can exaggerate this such that the third may lay down (with your example) 0.5mm on the outside. which with a circular model will generate 1mm oversize.

If I require accuracy, I will always extrude at 0.4mm. If I require strength I extrude 0.5mm. And I use those sizes because it makes modelling easier - I know that a 2mm wall will not have any gaps in the centre of it (with a 0.45mm extrusion there would be a 0.2mm gap).

Strangely most people create models in whole millimetres and sometimes, it matters!

Peter