Welcome! Log In Create A New Profile

Advanced

Wrong height and width

Posted by demetris 
Wrong height and width
January 25, 2013 02:05PM
Hello everyone,
this is my first post as a proud owner of Mendel90 smiling smiley

I have printed some really basic stuff so far and I wanted to start calibrating more. I tried to print an object (a small spanner). The result is really good for my current status but I noticed that the height and width are wrong. What is confusing to me, is that the height is shorter and the width is longer!

Check out the photos I am attaching to see. As I understand it, the reason for shorter height is the 'lack' of filament which could be fixed by adjusting the filament width in slicing software or STEPS_PER_MM from the firmware. However, this thought does no agree with the fact that the width is longer that it should, suggesting that I am extruding more filament than I should.

Please let me know what you think.

Thank you.
Attachments:
open | download - netfabb_head_width.png (30.3 KB)
open | download - netfabb_height.png (10.3 KB)
open | download - spanne_inside.png (267.1 KB)
open | download - spanner_height.png (191.8 KB)
open | download - spanner_width.png (281.4 KB)
Re: Wrong height and width
January 26, 2013 02:48PM
Hi Demetris,

Lack of filament won't affect the height. Assuming the first layer height is correct it should be "number of layers" times the "layer height". If the height of the model is not a multiple of the layer height then it won't be the same.

Try turning off extra outlines. I have come to realise that makes outlines too big and holes too small due to the fact that SF does not compensate for the fact the a single outline spreads symmetrically from the nozzle but a second outline is constrained to only spread in one direction.

If the size is wrong with a single outline then it looks like too much filament.


[www.hydraraptor.blogspot.com]
Re: Wrong height and width
January 27, 2013 05:47AM
Hi nophead,
I guess I still don't understand how the height and width of a line are formed. I printed the 0.5mm wall calibration cube as shown in the (really bad, sorry) attached picture, with some brim or else it wouldn't stick.

My settings are:
layer height: 0.3mm
perimeter extrusion width: 0.42mm

Cube Measurements:
Design height: 10mm
Measured height: 9.75mm
Design width: 22.5mm
Measured width: 22.61mm
Design wall thickness: 0.5mm
Measured wall thickness: 0.53mm

Now as I understand it, the measured wall thickness should be 0.42mm right? Since it is only one perimeter it should be the width of the perimeter in the slicer settings. Therefore, wall thickness is by 0.11mm more than it should.
On the other hand the height is off by 0.15mm right? Because closest you can get to is 9.9 with 33 layers.

So what is going on here? Sorry if my calculations are conceptually wrong but that is how I understand it at the moment!

Thanks!

Edited 2 time(s). Last edit at 01/27/2013 05:49AM by demetris.
Attachments:
open | download - 2013-01-27 12.28.42.jpg (61.8 KB)
Re: Wrong height and width
January 27, 2013 10:12AM
You say you are using s brim, which implies slic3r. I have never trusted the maths in that to get things the right size.

The height is formed by how much the Z axis moves and the width is determined by the flow rate versus the feed rate and the height.

For the height check what the Z values of the first and the last layer are in the gcode. They should be 0.3mm and 9.9mm. Also measure the thickness of the first layer. If it is printed at 0.3 but is only say 0.25 then adjust the Z axis in the firmware.


[www.hydraraptor.blogspot.com]
Re: Wrong height and width
January 27, 2013 05:01PM
I understand what you are saying, and I think I now understand better how it works. As you said I took some measurements and reduced the total height error to 0.05 ~ 0.01mm by lower the Z_HOME firmware parameter (i.e. raising the nozzle) by 0.1.

Now as I understand it, to reduce the width I will have to decrease the flow rate. I will report back with the results.
Sorry, only registered users may post in this forum.

Click here to login