pimples are still here ! #1037

Closed
jfpion opened this Issue Mar 5, 2013 · 23 comments

Projects

None yet

7 participants

@jfpion
jfpion commented Mar 5, 2013

there is less but still.....

granules
it's where the perimeters start and end

it will be a good thing to add the start and end inside the wall instead of just on the perimeter

thnak you

Owner
alexrj commented Mar 5, 2013

Do they appear in G-code? It looks like you need to calibrate your retraction settings.

jfpion commented Mar 5, 2013

there at the start end of the perimeters

there are here with different setting of retract from 1 to 4 mm retract length speed 20 mm/s

how to calibrate ? is there a method ?

the pb is with kissclicer same parameters there is no pimples (even in kisslicer doesn't have hoping

so it seem more there is a difference in the way to start/end the perimeter and possible improvement for slic3r

thank you

Owner
alexrj commented Mar 5, 2013

Do they appear in G-code?

jfpion commented Mar 5, 2013

no it doesn't appear on the Gcode but they are at the start stop of the perimeters

granules2

; generated by Slic3r 0.9.8 on 2013-03-04 at 15:35:09

; layer_height = 0.25
; perimeters = 2
; top_solid_layers = 3
; bottom_solid_layers = 3
; fill_density = 0.3
; perimeter_speed = 70
; infill_speed = 70
; travel_speed = 130
; scale = 1
; nozzle_diameter = 0.5
; filament_diameter = 3
; extrusion_multiplier = 0.95
; perimeters extrusion width = 0.80mm
; infill extrusion width = 0.80mm
; first layer extrusion width = 0.53mm
z lift 1mm
speed retract 20mm/s
2mm retract 0 extra length

look like there could be the use for a possibility to start/stop the perimeters inside like that

propal

thank you

Owner
alexrj commented Mar 5, 2013

That will cause a major blob because of extrudates overlapping.

Your picture shows clear calibration problems, and you confirmed there are no toolpath problems in G-code. However since you haven't posted any G-code, any config, any STL we can't really help you... Sorry.

@alexrj alexrj closed this Mar 5, 2013
jfpion commented Mar 5, 2013

you could have waited a bit before closing the issue i've got a very wobbly connexion i'm on the way to upload what you want

and you're wrong about blob that"s exactly what kisslicer does and there is no pimple and no blob
granules3

jfpion commented Mar 5, 2013

here are the gcode:
http://jean.francois.pion.free.fr/bayonet_container_slic3r.gcode
and the STL
http://jean.francois.pion.free.fr/bayonet_container_v1-2_20130304-15463-1560jtd-0.stl
http://jean.francois.pion.free.fr/bayonet_container_v1-2_20130304-15454-1hwybv1-0.stl

And despite this little pb I think slic3r is better than kisslicer (in fact that the only point I would like to see in slic3r which is in kisslicer )

tritol commented Mar 12, 2013

Same problem for me. Maybe it can be fixed with negative Extra length on restart (e.g -0.2 mm) but actual git version still generate invalid gcode as mentioned in issue #654. Should I post gcode examples there?

Owner
alexrj commented Mar 12, 2013

@tritol, if you think that issue is not fixed, please post your config.ini and a G-code file there so that I can check it.

Owner
alexrj commented Mar 12, 2013

@jfpion, a wiping move is just a non-extruding move inwards. It's different than starting the perimeter inside as in your first picture. Slic3r already does that, if you inspect the G-code closely.

jfpion commented Mar 13, 2013

Le 12/03/2013 16:05, Alessandro Ranellucci a écrit :

@jfpion https://github.com/jfpion, a wiping move is just a
non-extruding move inwards. It's different than starting the perimeter
inside as in your first picture. Slic3r already does that, if you
inspect the G-code closely.


Reply to this email directly or view it on GitHub
#1037 (comment).

an outward wipe mouvement after the restart of extrusion is not done and
it look like that's the cause of the small pimples

I think the sequence could be position the extruder slighty inward,
compensation of retractation, wipe movement outward to the start of
perimeter, start print,

to avoid the ooze making a small blob

thank you

JF Pion

_*
__Print "_Bonne année 2013*"

IF "La santé est bonne"

ALORS
" Continuer à profiter"*
_"Réaliser de beaux projets"
Sinon
"_Appeler SOS Médecins*"
Fin SI

While 2013 n'est pas fini
"Attendre 2014"
WEND

(ninos)

Des montages électroniques pour le modélisme
http://jean.francois.pion.free.fr
le site du vol électrique http://electrofly.free.fr/

Owner
alexrj commented Mar 13, 2013

Perimeters are already done starting from the inner loop, so no retraction compensation happens on the outside...

Did you try to lower your temperature? Raise your retraction speed?

Are you using a Bowden-tube printer?

jfpion commented Mar 13, 2013

Le 13/03/2013 20:10, Alessandro Ranellucci a écrit :

Perimeters are already done starting from the inner loop, so no
retraction compensation happens on the outside...

anyway it is something connected with the start stop of the perimeter
because the repartition is totally different if I ask random start or not

Did you try to lower your temperature?

yes and the conclusion is too low is not good for the nozzle screw in
the peek, the extruder push hard enough (no need to increase the
démultiplication and the filament is not slipping) at the end the nozzle
pop out the peek, the ptfe go out with cold abs, i had to got a new hot
end (was a jhead mark5 0.5mm) so no new test for now

Raise your retraction speed?

20mm/s seems enough ?

Are you using a Bowden-tube printer?

no 3mm filament wade extruder direct feed to the hotend

what make me to think there is something to improve in the gcode is the
same part sliced whith kisslicer had not these pimples

the only difference is kisslicer doesn't apply Z lift

may be the Z lift could be done with a combination of move lift AND
inward to avoid this kind of pimple ?

I will try as soon as as possible the results of some try I want to do
because i'm starting to think the truth is elswhere...

(i'm having some over shooting on the perimeters when printing hexagone,
the overshoot is consistent on all the angles not pnly the ones aligned
on x or y so as this oversoot is not in the gcode the answer could be in
marlin with the built in acceleration feature. the backlash test i've
done is quite good and the baclash could not explain the overshoot angles

I will print the same part with a rather slow speed for the external
perimeter to see if there is difference as soon as I got an anothe jhead

JF Pion

_*
__Print "_Bonne année 2013*"

IF "La santé est bonne"

ALORS
" Continuer à profiter"*
_"Réaliser de beaux projets"
Sinon
"_Appeler SOS Médecins*"
Fin SI

While 2013 n'est pas fini
"Attendre 2014"
WEND

(ninos)

Des montages électroniques pour le modélisme
http://jean.francois.pion.free.fr
le site du vol électrique http://electrofly.free.fr/

nophead commented May 24, 2013

Adrian Bowyer blogged starting in the inside years ago and I have been doing that for years also. I start an island with an infill line. You still get a seam (but not a blob) because filament has a minimum bend radius.

Owner
alexrj commented May 24, 2013

@nophead, are you able to find links to such blog posts?

I already forwarded prior art from Slic3r repository to the Electronics Frontier Foundation to help stop the patent. Anything you have about that would be so precious.

jfpion commented May 24, 2013

If you want to use my prior intervention to stop that, i totally agree
if i can do somthing else to help, please, say it

the beginning of the fame to be robbed of an idea ;-)

nophead commented May 24, 2013

I haven't found any links yet so I have emailed Adrian.

On 24 May 2013 16:47, jfpion notifications@github.com wrote:

If you want to use my prior intervention to stop that, i totally agree
if i can do somthing else to help, please, say it

the beginning of the fame to be robbed of an idea ;-)


Reply to this email directly or view it on GitHubhttps://github.com/alexrj/Slic3r/issues/1037#issuecomment-18413039
.

nophead commented Sep 20, 2016

I finally managed to get in touch with Adrian and from the date in the Java host source code I was able to find the blog post. http://blog.reprap.org/2009/07/build-quality.html

Contributor
bubnikv commented Sep 23, 2016

Cool, so now we have a prior art and we may experiment with the idea. Now, will this be useful? How to handle multi-perimeter prints? Will it work for inner perimeters first? Would it require a special way to plan the inner perimeters, so there is a space for the inside/outside kink after the inner perimeter is printed?

Contributor
VanessaE commented Oct 4, 2016

It would "seam" ( 😝 ) that we don't need to worry about the patent anymore: http://www.cafc.uscourts.gov/sites/default/files/opinions-orders/15-1769.Opinion.9-28-2016.1.PDF

(Short version: software patents are unconstitutional in the US, by way of the 1st amendment)

Collaborator

That's going up to SCOTUS.

"Nothing unreal exists." - Kiri-kin-tha's First Law of Metaphysics.

On Tue, Oct 4, 2016 at 6:35 PM, Vanessa Ezekowitz notifications@github.com
wrote:

It would "seam" ( 😝 ) that we don't need to worry about the patent
anymore: http://www.cafc.uscourts.gov/sites/default/files/opinions-
orders/15-1769.Opinion.9-28-2016.1.PDF

(Short version: software patents are unconstitutional in the US, by way of
the 1st amendment)


You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#1037 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAB8CsdcOlEznL7pGwpznHqdN1xP2ezoks5qwuLQgaJpZM4AefGX
.

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