Huge E distances at random #1691

Closed
T3P3 opened this Issue Jan 10, 2014 · 10 comments

Comments

Projects
None yet
3 participants
@T3P3

T3P3 commented Jan 10, 2014

GCode being generated while slicing a multi-material amf contain a few lines which call for "huge" extruder distances. The file in question has 4 parts but the error happens to be while T0 is selected.

I am using Slic3r 10.0.0RC2, config is on the dropbox share along with the stls, amf and generated gcode (I scaled the amf to 80% before slicing).
https://www.dropbox.com/sh/rehmsd3x53kb624/0vtqRRFRAq

example of GCode in question (with my comments added):

G1 X69.429 Y115.002 F3600.000
G1 X69.409 Y114.982 E0.00100 F1200 ;normal amount to be extruded for short move
G1 F3600.000 E-2.00000
G1 X74.349 Y116.419 F3600.000
G1 E2.00000 F3600.000
G1 X75.240 Y117.310 E9.98607 F1800 ;~9000x too much?
G1 F3600.000 E-2.00000
G1 X71.205 Y113.050 F3600.000
G1 E2.00000 F3600.000
G1 X65.673 Y107.518 E62.00709 F1800 ;62mm
G1 F3600.000 E-2.00000
G1 X78.904 Y71.728 F3600.000
G1 E2.00000 F3600.000
G1 X107.839 Y100.663 E324.35853 F1800 ;324 mm!
G1 F3600.000 E-2.00000
G1 X111.579 Y114.220 F3600.000
G1 E2.00000 F3600.000
G1 X115.072 Y117.713 E39.14928 F1800 ;39mm

This occurs around line number 20K of the particular file, overall only 7 times in this file is the Extruder distance weird, 6 times at this point and 1 ~10K lines later so its not happening very often (enough to ruin the print though!).

I have not had a chance to check with other object, at least one other 4 material print sliced fine but this print when scaled to 100% has the same issues.

Cheers

Tony

@alexrj

This comment has been minimized.

Show comment
Hide comment
@alexrj

alexrj Jan 10, 2014

Member

Hello Tony!
That sounds like some overflow. I'll have a look at it as soon as possible.

Member

alexrj commented Jan 10, 2014

Hello Tony!
That sounds like some overflow. I'll have a look at it as soon as possible.

@T3P3

This comment has been minimized.

Show comment
Hide comment
@T3P3

T3P3 Jan 10, 2014

Cheers

Alessandro!
On 10 Jan 2014 16:57, "Alessandro Ranellucci" notifications@github.com
wrote:

Hello Tony!
That sounds like some overflow. I'll have a look at it as soon as possible.


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32044720
.

T3P3 commented Jan 10, 2014

Cheers

Alessandro!
On 10 Jan 2014 16:57, "Alessandro Ranellucci" notifications@github.com
wrote:

Hello Tony!
That sounds like some overflow. I'll have a look at it as soon as possible.


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32044720
.

@alexrj

This comment has been minimized.

Show comment
Hide comment
@alexrj

alexrj Jan 10, 2014

Member

So, it's not an overflow but it looks like some part of the retraction or tool change logic doesn't take the use_relative_e_distances option into account. Not many people use that option, so that's why this bug slipped in. I'm trying to write a regression test before working on the actual fix.

Member

alexrj commented Jan 10, 2014

So, it's not an overflow but it looks like some part of the retraction or tool change logic doesn't take the use_relative_e_distances option into account. Not many people use that option, so that's why this bug slipped in. I'm trying to write a regression test before working on the actual fix.

@T3P3

This comment has been minimized.

Show comment
Hide comment
@T3P3

T3P3 Jan 10, 2014

Thanks! The reprap firmware uses relative E distances as a default.
On 10 Jan 2014 23:21, "Alessandro Ranellucci" notifications@github.com
wrote:

So, it's not an overflow but it looks like some part of the retraction or
tool change logic doesn't take the use_relative_e_distances option into
account. Not many people use that option, so that's why this bug slipped
in. I'm trying to write a regression test before working on the actual fix.


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32077083
.

T3P3 commented Jan 10, 2014

Thanks! The reprap firmware uses relative E distances as a default.
On 10 Jan 2014 23:21, "Alessandro Ranellucci" notifications@github.com
wrote:

So, it's not an overflow but it looks like some part of the retraction or
tool change logic doesn't take the use_relative_e_distances option into
account. Not many people use that option, so that's why this bug slipped
in. I'm trying to write a regression test before working on the actual fix.


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32077083
.

@alexrj

This comment has been minimized.

Show comment
Hide comment
@alexrj

alexrj Jan 11, 2014

Member

Okay, it's not a bug.
You have the following in your config:

top_infill_extrusion_width = 90

which means 90mm, and this is why it's trying to extrude that much.

However, the difficulty of finding such error means that some action needs to be taken as this is terrible user experience. I think I'll put a hard limit on extrusion widths even when specified manually.

Member

alexrj commented Jan 11, 2014

Okay, it's not a bug.
You have the following in your config:

top_infill_extrusion_width = 90

which means 90mm, and this is why it's trying to extrude that much.

However, the difficulty of finding such error means that some action needs to be taken as this is terrible user experience. I think I'll put a hard limit on extrusion widths even when specified manually.

@alexrj

This comment has been minimized.

Show comment
Hide comment
@alexrj

alexrj Jan 11, 2014

Member

Done. A hard limit is now in place.

Member

alexrj commented Jan 11, 2014

Done. A hard limit is now in place.

@alexrj alexrj closed this Jan 11, 2014

alexrj added a commit that referenced this issue Jan 11, 2014

Put a hard limit on manually configured extrusion widths (10 * the la…
…rgest nozzle diameter configured) to prevent confusion when a bad value is entered. #1691

alexrj added a commit that referenced this issue Jan 11, 2014

Put a hard limit on manually configured extrusion widths (10 * the la…
…rgest nozzle diameter configured) to prevent confusion when a bad value is entered. #1691

Conflicts:

	lib/Slic3r/Config.pm
@T3P3

This comment has been minimized.

Show comment
Hide comment
@T3P3

T3P3 Jan 11, 2014

Ahhh sorry it was 0.9 I must have messed it up when trying different
settings.

Thanks for sorting a hard limit for foolish users like me!
On 11 Jan 2014 13:31, "Alessandro Ranellucci" notifications@github.com
wrote:

Done. A hard limit is now in place.


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32095969
.

T3P3 commented Jan 11, 2014

Ahhh sorry it was 0.9 I must have messed it up when trying different
settings.

Thanks for sorting a hard limit for foolish users like me!
On 11 Jan 2014 13:31, "Alessandro Ranellucci" notifications@github.com
wrote:

Done. A hard limit is now in place.


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32095969
.

@hroncok

This comment has been minimized.

Show comment
Hide comment
@hroncok

hroncok Jan 11, 2014

Member

Wouldn't a warning be better than limit?

Member

hroncok commented Jan 11, 2014

Wouldn't a warning be better than limit?

@alexrj

This comment has been minimized.

Show comment
Hide comment
@alexrj

alexrj Jan 11, 2014

Member

@hroncok, probably yes but I don't know what would be a good method for have Slic3r warn the user (in GUI mode). A dialog that pops up every time asking for confirmation doesn't sound very comfortable. Also, if someone is able to extrude with a width equal to 10 * nozzle diameter I definitely want them to share pictures with me ;)

Member

alexrj commented Jan 11, 2014

@hroncok, probably yes but I don't know what would be a good method for have Slic3r warn the user (in GUI mode). A dialog that pops up every time asking for confirmation doesn't sound very comfortable. Also, if someone is able to extrude with a width equal to 10 * nozzle diameter I definitely want them to share pictures with me ;)

@hroncok

This comment has been minimized.

Show comment
Hide comment
@hroncok

hroncok Jan 11, 2014

Member

Makes sense, once it is possible, they will file another issue :)
Dne 11.1.2014 15:21 "Alessandro Ranellucci" notifications@github.com
napsal(a):

@hroncok https://github.com/hroncok, probably yes but I don't know what
would be a good method for have Slic3r warn the user (in GUI mode). A
dialog that pops up every time asking for confirmation doesn't sound very
comfortable. Also, if someone is able to extrude with a width equal to 10 *
nozzle diameter I definitely want them to share pictures with me ;)


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32097128
.

Member

hroncok commented Jan 11, 2014

Makes sense, once it is possible, they will file another issue :)
Dne 11.1.2014 15:21 "Alessandro Ranellucci" notifications@github.com
napsal(a):

@hroncok https://github.com/hroncok, probably yes but I don't know what
would be a good method for have Slic3r warn the user (in GUI mode). A
dialog that pops up every time asking for confirmation doesn't sound very
comfortable. Also, if someone is able to extrude with a width equal to 10 *
nozzle diameter I definitely want them to share pictures with me ;)


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1691#issuecomment-32097128
.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment