What is causing this weird behavior? (Skeinforge issue?)
Dave M.
in its own thread.
Anyhow, I really wanted to get the calibration of the machine down so
that the extrusion speeds result in the right filament thickness. I
think I am doing it correctly, but I'm seeing strange behavior that I
can't explain:
First, when I print the single wall calibration piece, after each
completion of the square, the X axis jogs back and forth before laying
the next layer. This severely messes everything up. I posted a
video on youtube here so you can see what I'm talking about.
http://www.youtube.com/watch?v=Qln2ag4SsvY. It didn't do this on the
first 5 or 6 test pieces, though! I just deleted all of my Skeinforge
filter settings and started over from scratch, manually inputting the
number I had been recording. Same problem!!!
Next, when I changed my layer thickness from 0.4 to 0.36, the single
wall test piece ended up getting printed with a triple wall
thickness. I can't explain this behavior, either. I am starting a
new print now at 0.36 to see if it repeats this strangeness.
Also, I thought that maybe the warping I am experiencing is not only
due to the lack of a heated raft, but also because the nozzle temp is
too high. Does everyone advise that I finally sit down and attempt to
update the calibration table for the 1mm thermistor? Right now I'm
running at 250C, though this morning I found that 240C extrudes as
well. The problem is that the raft doesn't adhere in the beginning
and goops up. I'll post a pic in my blog later today.
Finally, as far as calibration goes, I've been using 1.6 for these
settings:
carve:extrusion width over thickness
inset: extrusion perimeter width over thickness
speed: extrusion diameter over thickness
Do you guys recommend fixing those values at 1.6 and then dialing in
the extrusion speed, travel speed, and layer height? Or should I
start with a slightly lower number, like 1.45 and redo the
calculations?
Thanks for your help!!!
Cathal Garvey
I can only suggest anything heat-wise, but I personally use preheat @ 240C, printing at around 227/230 and have very few warping issues. I print my raft low down at 233 and it adheres really well, too. However, at 220 my extruder will strip/break.
--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To post to this group, send email to make...@googlegroups.com.
To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.
Cliff Biffle
> Next, when I changed my layer thickness from 0.4 to 0.36, the single
> wall test piece ended up getting printed with a triple wall
> thickness.
Did you change your extrusion width/diameter over layer thickness
accordingly? Since they're ratios of the layer thickness, changing
the layer thickness means you have to adjust them.
Further evidence that Skeinforge was written by martians. :-)
> Do you guys recommend fixing those values at 1.6 and then dialing in
> the extrusion speed, travel speed, and layer height? Or should I
> start with a slightly lower number, like 1.45 and redo the
> calculations?
I recommend fixing the feedrate, layer height, and PWM, and then
setting those from empirical measurements. If you can read Python the
difference between the settings is pretty clear: extrusion diameter is
for plastic extruded into space, to calculate the volumetric
throughput; extrusion *width* is when fusing a layer, and can be
derived from your single-wall test piece.
-Cliff L. Biffle
Dave M.
I actually didn't change any other parameters, because I am trying to
figure out how all of the settings are related. I thought that the
layers might gets squished a little more since the layer height
determines the Z up move for each layer (right?), but I did not expect
it to do multiple passes of the wall at the exact same Z height. I'll
keep it at 0.4 for now, but in all of my parts I'd like the outer
walls to be a little less rough (i.e. more mushed together).
> I recommend fixing the feedrate, layer height, and PWM, and then
> setting those from empirical measurements.
Sorry, I don't understand what you mean by this. Are you saying that
I should figure out the ideal values for layer height, PWM, and travel
feedrate, and *then* fix them there permanently? Or are you saying I
should just set them to whatever I want, and then adjust other
parameters accordingly? If the latter, then I assume I have to tinker
with all of the extrusion width settings that I had previously set to
1.6?
Thanks!
On Jan 15, 8:42 am, Cliff Biffle <cbif...@gmail.com> wrote:
Cliff Biffle
> Sorry, I don't understand what you mean by this. Are you saying that
> I should figure out the ideal values for layer height, PWM, and travel
> feedrate, and *then* fix them there permanently? Or are you saying I
> should just set them to whatever I want, and then adjust other
> parameters accordingly?
The latter, yes.
> If the latter, then I assume I have to tinker
> with all of the extrusion width settings that I had previously set to
> 1.6?
Yes. Other people disagree with me on this, but I haven't gotten
great results by picking a number and setting all those settings to
it. I read the Skeinforge code and did some math to derive my
settings, and they work great -- even fills, solid non-bloopy curves,
etc.
I should really post this procedure on the web so I can quit typing it
here. :-) But essentially what I would recommend is:
1. Choose the temperature that lets you extrude easily and quickly,
without too much smoking or popping (or catching fire or...).
2. Choose a layer height. I use the default of 0.36. Within
reasonable bounds you can pick whatever you're comfortable with.
You're correct that this controls the Z movements; it also figures
into the flowrate calculations.
3. Choose a PWM value. I personally use 230 -- it lets me extrude
finer features than 255 by reducing the flowrate. My procedure for
this consisted of extruding 10 seconds of filament at various PWM
values, regressing the relationship between the length and PWM (it
seems roughly linear down to 200), and throwing all that away and
picking the middle one arbitrarily. :-)
4. Choose a feedrate (the speed the head moves while printing). I
stuck with the default -- too fast, and you get tears and gaps in the
model; too slow, and you get bored. :-)
Given those independent(ish) variables, you can empirically derive the
Skeinforge parameters:
5. Extrude some filament into space (like the test extrusion). Don't
pull it or otherwise change its shape. Measure its diameter with
calipers. This is your "extrusion diameter."
6. Print a single-walled test object. Measure the width of the wall
(on a layer well above the raft -- the first layer or two are often
aberrant). This is your "extrusion width."
Divide these values by your chosen layer thickness to set the
parameters. In my case, the extrusion diameter/layer thickness is in
the neighborhood of 1.6 (1.6111 iirc, I'm not at my workshop) but the
extrusion width/thickness wound up slightly lower, around 1.5. If I
changed any of the independent variables in steps 1-4 I would need to
recompute these.
The purely empirical settings may cause slightly sparse fills: you
actually want adjacent extrusions to overlap somewhat. I did some
over-complicated volumetric computations based on my measured flowrate
and the degree of "squish" I wanted, but honestly I think you can just
nudge the "extrusion width over thickness" parameter down slightly to
fix it. (Remember: this parameter doesn't *control* the extrusion
width -- it *describes* it, to Skeinforge, so it can decide how much
room to leave.)
The "extrusion perimeter width" is the extrusion width used on the
perimeter. You can tweak it away from 1.0 if you want the perimeter
to be more or less solid. I leave it unchanged.
-Cliff L. Biffle
Dave M.
the list of items and work through them next week. Exciting stuff!
On Jan 15, 11:06 am, Cliff Biffle <cbif...@gmail.com> wrote:
Dave M.
have no idea why my machine is doing this, but it keeps laying down
layers and then making dry passes around the perimeter. I believe
that I can see them in Skeinviewer -- they come up as gray arrows. So
now I'm playing with the settings to see what might make that
weirdness go away.
What I don't get is that even when I go back to default settings (by
deleting the .skeinforge folder in my user folder), it still does
this, but it didn't do this before!!!
Does anyone know what setting will make that behavior go away? It's
like it's doing a cooling pass, but I have cooling disabled.
Andrew Plumb
Threads/..." settings in Raft. Those are what cause the orbiting
behaviour.
Andrew.
> --
> You received this message because you are subscribed to the Google
> Groups "MakerBot Operators" group.
> To post to this group, send email to make...@googlegroups.com.
> To unsubscribe from this group, send email to makerbot+u...@googlegroups.com
> .
> For more options, visit this group at http://groups.google.com/group/makerbot?hl=en
> .
>
>
--
"The future is already here. It's just not very evenly distributed"
-- William Gibson
Dave M.
On Jan 15, 12:35 pm, Andrew Plumb <and...@plumb.org> wrote:
> Zero-out all the "Time Before Raft/First Layer Outline/Next
> Threads/..." settings in Raft. Those are what cause the orbiting
> behaviour.
>
> Andrew.
>
> On 15-Jan-10, at 3:28 PM, Dave M. wrote:
>
>
>
> > I started to work on calibration again, and it's really screwy. I
> > have no idea why my machine is doing this, but it keeps laying down
> > layers and then making dry passes around the perimeter. I believe
> > that I can see them in Skeinviewer -- they come up as gray arrows. So
> > now I'm playing with the settings to see what might make that
> > weirdness go away.
>
> > What I don't get is that even when I go back to default settings (by
> > deleting the .skeinforge folder in my user folder), it still does
> > this, but it didn't do this before!!!
>
> > Does anyone know what setting will make that behavior go away? It's
> > like it's doing a cooling pass, but I have cooling disabled.
> > --
> > You received this message because you are subscribed to the Google
> > Groups "MakerBot Operators" group.
> > To post to this group, send email to make...@googlegroups.com.
> > To unsubscribe from this group, send email to makerbot+u...@googlegroups.com
> > .
> > For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en
Dave M.
and they are all zero:
Format is tab separated preferences.
Activate Raft: True
Add Raft, Elevate Nozzle, Orbit and Set Altitude: True
Base Infill Density (ratio): 0.5
Base Layer Thickness over Layer Thickness: 1.0
Base Layers (integer): 1
Base Nozzle Lift over Half Base Layer Thickness (ratio): 0.4
Bottom Altitude: 0.1
Open File to be Rafted
Infill Overhang (ratio): 0.1
Interface Infill Density (ratio): 0.5
Interface Layer Thickness over Layer Thickness: 0.8
Interface Layers (integer): 1
Interface Nozzle Lift over Half Interface Layer Thickness (ratio): 0.5
Operating Nozzle Lift over Half Layer Thickness (ratio): 1.0
Raft Outset Radius over Extrusion Width (ratio): 7.5
Support Inset over Perimeter Extrusion Width (ratio): 0.0
No Support Material True
Support Material Everywhere False
Support Material on Exterior Only False
Support Minimum Angle (degrees): 60.0
Temperature Change Time Before Raft (seconds): 0.0
Temperature Change Time Before First Layer Outline (seconds): 0.0
Temperature Change Time Before Next Threads (seconds): 0.0
Temperature Change Time Before Support Layers (seconds): 0.0
Temperature Change Time Before Supported Layers (seconds): 0.0
Temperature of Raft (Celcius): 230.0
Temperature of Shape First Layer Outline (Celcius): 215.0
Temperature of Shape First Layer Within (Celcius): 215.0
Temperature of Shape Next Layers (Celcius): 220.0
Temperature of Support Layers (Celcius): 220.0
Temperature of Supported Layers (Celcius): 220.0
windowPositionRaft Preferences 0+0
On Jan 15, 12:35 pm, Andrew Plumb <and...@plumb.org> wrote:
> Zero-out all the "Time Before Raft/First Layer Outline/Next
> Threads/..." settings in Raft. Those are what cause the orbiting
> behaviour.
>
> Andrew.
>
> On 15-Jan-10, at 3:28 PM, Dave M. wrote:
>
>
>
> > I started to work on calibration again, and it's really screwy. I
> > have no idea why my machine is doing this, but it keeps laying down
> > layers and then making dry passes around the perimeter. I believe
> > that I can see them in Skeinviewer -- they come up as gray arrows. So
> > now I'm playing with the settings to see what might make that
> > weirdness go away.
>
> > What I don't get is that even when I go back to default settings (by
> > deleting the .skeinforge folder in my user folder), it still does
> > this, but it didn't do this before!!!
>
> > Does anyone know what setting will make that behavior go away? It's
> > like it's doing a cooling pass, but I have cooling disabled.
> > --
> > You received this message because you are subscribed to the Google
> > Groups "MakerBot Operators" group.
> > To post to this group, send email to make...@googlegroups.com.
> > To unsubscribe from this group, send email to makerbot+u...@googlegroups.com
> > .
> > For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en
Andrew Plumb
never turned it on.
A.
> For more options, visit this group at http://groups.google.com/group/makerbot?hl=en
Cliff Biffle
> Do you have Cool enabled? That might cause orbiting behaviour; I've never
> turned it on.
Definitely, this was my first thought. In Skeinforge 0006 cool causes
it to rub the outer perimeter (instead of orbiting outside it as in
newer versions).
-Cliff L. Biffle
Dave M.
there, so I'll zero them out just in case. This is so weird.
Dave M.
Skeinview shows that the weird orbits are no longer occurring, but now
I'm a bit hesitant to go with a newer version. I'll have to take this
very slowly, I think.
Dave M.
the old version of Skeinforge and can't get the weird orbiting or
jiggles to stop. I loaded the latest Skeinforge, but even though it
doesn't exhibit the same weird behavior (according to SkeinView), it
is also so drastically different that I haven't found the time to mess
with the gcode export so that it won't crash my Cupcake.
Here's a link to my SkeinView output. You'll see that several layers
have weird light gray paths, which is where the stage moves without
actually extruding anything.
http://farm5.static.flickr.com/4071/4292340492_ff208c1c84_o.gif
Any info would be greatly appreciated! I've tried all of the
suggestions on this thread so far, and none of them have fixed the
problem. It's very odd that Skeinforge used to output gcode for my
Cupcake that worked! Once again, I'm printing the single-walled test
piece from Thingiverse.
Thanks!
Steven Dick
Sorry, I need to bump this post again! I have tried everything with
the old version of Skeinforge and can't get the weird orbiting or
jiggles to stop.
I've seen both the extrudeless or mostly extrudeless orbiting and the jiggling when you have layers that skeinforge wants to make solid that are a fraction of a wall thickness.
For example, if your wall thickness is set to 0.35 (the default I think), and your part part has a thin wall that is 0.875 (or 2.5 walls) thick, then it will draw the inner wall and the outer wall, and then make an orbit where it extrudes half as much (or moves twice as fast) to fill in the gap.
If your thin wall is instead 3.5 thick, it'll make an inner wall and an outer wall, and then jiggle for the 1.5 wall thickness difference to fill the gap.
These fractional thicknesses are really bad if you don't have your extruder calibrated, because when it tries to fill the gap, it will either put out far too much material and blob badly and run into those blobs on the very next pass, or it will not put out enough material and the walls will be full of gaps and not solid enough to have any strength.
The whistle is especially sensitive to this, as all the walls are a small non-multiple of 0.35 and also fairly tall, so lots of opportunities for blobbing. The length of the wall is also long enough that the blobs tend to harden before the nozzle passes over them a second time, so rather than passing through them and maybe picking the blob up, it knocks the part off the paltform or worse.
I'm not sure why you don't like the orbits (I find they frequently fix blobs and smooth out the surface), but if your steppers and belts are not right, the jiggles would probably cause all kinds of problems.
Dave M.
sense. It is ironic that I'm experiencing this behavior because my
machine isn't calibrated, but the reason why I'm printing the single-
walled test piece is because I'm trying to calibrate the machine! :)
The page for the test piece (http://www.thingiverse.com/thing:1549)
states that the walls should each be 0.5mm thick for the 0.5mm
nozzle. From previous posts, I thought I was supposed to pick a wall
thickness that I was happy with and dial in the settings to get it.
But if the piece here is supposed to be 0.5mm, then obviously I should
shoot for that or every part I make will come out wrong. The creator
says that the walls are actually 0.6, 0.5, 0.4, and 0.3mm for dialing
in the 0.35mm nozzle, but I don't understand how this could work --
how can it end up being 0.5mm on each side for a 0.5mm nozzle if
Skeinforge doesn't know this?
Hopefully, I'll have time to try this tomorrow. I haven't fired up
the Cupcake since last Friday. :( Thanks again for the reply!
On Jan 20, 10:34 pm, Steven Dick <kg4...@gmail.com> wrote: