1 Topic by fischelbyxa

  • fischelbyxa
  • fischelbyxa
  • Member
  • Offline
  • Registered: 2013-04-02
  • Posts: 48

Topic: Defective print - Printer or Slic3r?

Two images, the first is a 2D export from Sketchup showing what I was hoping for, the second a photograph of the actual result (with captions).

The defects:

1. The "gills" are slanted at 45 degrees. They are supposed to be solid, but there's a gap in all of them. It's like if the G-code assumed the walls would meet in the middle. Could this be flow rate (the walls should be thicker, but due to too low flow rate, there're not)? Or is it a problem in Slic3r (a plain old bug)? Or?

2. The two holes are not round. Is this an error in the printer setup (X and Y axis are not moving correct distances compared to each other)? Or?

3. One of the gills got one layer of fill, there should be none. Bug in Slic3r? Or?

4. The join between a floor and a wall leaves a massive gap. Slic3r? Misaligned printer? Or?

5. The overall measurements are slightly off: Length should be 120 mm, it is 120.20 mm. Width should be 70 mm, it is 70.84 mm. X and Y misalignment? Or?

Post's attachments

PiDolly-11.png
PiDolly-11.png 103.64 kb, file has never been downloaded. 

PiDolly-defects.png
PiDolly-defects.png 778.83 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

2 Reply by adrian (edited by adrian 2013-06-09 14:46:43)

  • adrian
  • Member
  • Offline
  • From: Canberra, Australia
  • Registered: 2013-01-15
  • Posts: 1,623

Re: Defective print - Printer or Slic3r?

Your Y-motor belt needs tightening. thats the cause of the oval holes. You may also have problems with your belt tension being uniform, but start with the y-motor.

The one-layer fill could be slic3r, could be the model. Run it through cloud netfabb and see if the problem persists.

the gaps are indeed flow-rate.. have you calibrated your extruder steps-per-mm and flow-rate as per the wiki ?

the off measurements are also due to steps-per-mm, and can also be due to misalignment and backlash in the x/y planes.

First stop i'd say is bung it through cloud netfabb, check your y-motor is tight without bending the link-rod, do a full extruder calibration of steps-per-mm and flow-rate using a test cube, and then reprint...

3 Reply by fischelbyxa

  • fischelbyxa
  • fischelbyxa
  • Member
  • Offline
  • Registered: 2013-04-02
  • Posts: 48

Re: Defective print - Printer or Slic3r?

adrian wrote:

Your Y-motor belt needs tightening. thats the cause of the oval holes. You may also have problems with your belt tension being uniform, but start with the y-motor.

Just out of curiosity: What makes you target the Y axis and not the X axis?

adrian wrote:

The one-layer fill could be slic3r, could be the model. Run it through cloud netfabb and see if the problem persists.

Already did. The results are from a netfabb checked STL-file. I tried my own some time ago, that print had a lot of other problems, but the one-layer fill wasn't there.

adrian wrote:

the off measurements are also due to steps-per-mm, and can also be due to misalignment and backlash in the x/y planes.

You kind of thought that would be correct when you get it from Solidoodle. Print straight out of the box and all...

4 Reply by ronsii

  • ronsii
  • Member
  • Offline
  • From: PNW
  • Registered: 2012-10-28
  • Posts: 1,723

Re: Defective print - Printer or Slic3r?

fischelbyxa wrote:
adrian wrote:

Your Y-motor belt needs tightening. thats the cause of the oval holes. You may also have problems with your belt tension being uniform, but start with the y-motor.

Just out of curiosity: What makes you target the Y axis and not the X axis?

The y axis is usually just more likely to have issues with multiple belts and shafts you have a lot more tension problems and such especially with environment changes.

adrian wrote:

The one-layer fill could be slic3r, could be the model. Run it through cloud netfabb and see if the problem persists.

Already did. The results are from a netfabb checked STL-file. I tried my own some time ago, that print had a lot of other problems, but the one-layer fill wasn't there.

I have done a few models in sketckup and netfab and some still ended up with problems, I have been using freecad lately and once you get the hang of what not to do smile it seems to slice better than similar models from sketchup.

adrian wrote:

the off measurements are also due to steps-per-mm, and can also be due to misalignment and backlash in the x/y planes.

You kind of thought that would be correct when you get it from Solidoodle. Print straight out of the box and all...

Well... they didn't say 'prints accurately out of the box' wink

5 Reply by RavensCrest

  • RavensCrest
  • Member
  • Offline
  • Registered: 2013-01-12
  • Posts: 269

Re: Defective print - Printer or Slic3r?

fischelbyxa wrote:

1. and 4.

It might be too thin to extrud infill for that area. Can you post a gcode's image and check to see if plastic is extruded in that area or not. (if it is... it might be retraction)

fischelbyxa wrote:

3. One of the gills got one layer of fill, there should be none. Bug in Slic3r? Or?

looks like brim to me.. model with holes will have random brim created for it.. once again. check your gcode's graphical display, and display first layer and see if it's there or not with brim on and off.

Solidoodle2 with Ceramic tile heated bed http://www.soliforum.com/topic/2544/my- … eated-bed/
"1kg should last for an while" is a lie!

6 Reply by Briggs

  • Briggs
  • Member
  • Offline
  • From: Minnesota
  • Registered: 2013-03-20
  • Posts: 226

Re: Defective print - Printer or Slic3r?

I'm with Adrian on this one, it's most likely your y axis. Why not the x? Because it is unlikely to have slop. Your long y axis belts probably don't have slop either. Your little jack shaft drive belt.....well....now that probably has all your slop. Just take your fingers and move the y stepper. Look at the jack shaft when you do it. Do you see the slop??

As for belt tension on the long y axis belts, that can affect your binding of the x carriage, and this needs to be evenly tensioned for this reason.

Also, don't overlook loose grub (set) screws and  be sure your rods are clean and greased.

7 Reply by fischelbyxa

  • fischelbyxa
  • fischelbyxa
  • Member
  • Offline
  • Registered: 2013-04-02
  • Posts: 48

Re: Defective print - Printer or Slic3r?

Wouldn't it be nice to have a way know when the tension is right? One could use a Newton meter to measure the force required to pull the belt X mm.

In Sweden we have the term "lagom" meaning "just right." It's a bit like Goldielock, not too much, not too little. As a Swden, I have the "lagom" concept ingrained in my DNA since time immemorial. But when calibrating the tension of e.g. drive belt, "lagom" isn't the best way to describe the end result...


Lagom: http://en.wikipedia.org/wiki/Lagom

8 Reply by adrian

  • adrian
  • Member
  • Offline
  • From: Canberra, Australia
  • Registered: 2013-01-15
  • Posts: 1,623

Re: Defective print - Printer or Slic3r?

Not scientific, but it at least ensures you have them evenly set: http://www.thingiverse.com/thing:38952

9 Reply by fischelbyxa

  • fischelbyxa
  • fischelbyxa
  • Member
  • Offline
  • Registered: 2013-04-02
  • Posts: 48

Re: Defective print - Printer or Slic3r?

adrian wrote:

Not scientific, but it at least ensures you have them evenly set: http://www.thingiverse.com/thing:38952

Perfect! A lot better than not too tight, not too lose!

Thank you, Adrian!

10 Reply by adrian

  • adrian
  • Member
  • Offline
  • From: Canberra, Australia
  • Registered: 2013-01-15
  • Posts: 1,623

Re: Defective print - Printer or Slic3r?

fischelbyxa wrote:
adrian wrote:

Not scientific, but it at least ensures you have them evenly set: http://www.thingiverse.com/thing:38952

Perfect! A lot better than not too tight, not too lose!

Thank you, Adrian!

No problems.

I also meant to tell you, re your model accuracy, be sure to tick 'External Perimeters First' in Slic3r config - this will draw from the outside inwards instead of the default inwards-out and helps improve accuracy a lot as it accounts for 'ooze' in the plastics much better etc.