Mosaic Questions...
122 views
Skip to first unread message
electronbee
Aug 19, 2011, 4:16:28 PM8/19/11
to MakerGear - Make Today, Change Tomorrow
I did not see a link to a manual for the Mosaic on the site...
Any ways, I am seriously considering the Mosaic as all i really want
to do is print and I am not too concerned with tinkering with the
machine itself. Plus, I want a plug-n-play product so I have as little
to do set-up wise as possible. That being said:
1) Is a calibration print required?
2) Can calibration at a later date be accomplished if it goes "out"?
3) Is the only print head available for 1.75mm ABS and not 3mm? I only
ask as the 3mm material selection seems to be greater.
4) Any videos of it in action?
Thanks!
Any ways, I am seriously considering the Mosaic as all i really want
to do is print and I am not too concerned with tinkering with the
machine itself. Plus, I want a plug-n-play product so I have as little
to do set-up wise as possible. That being said:
1) Is a calibration print required?
2) Can calibration at a later date be accomplished if it goes "out"?
3) Is the only print head available for 1.75mm ABS and not 3mm? I only
ask as the 3mm material selection seems to be greater.
4) Any videos of it in action?
Thanks!
Luke
Aug 20, 2011, 4:37:46 PM8/20/11
to MakerGear - Make Today, Change Tomorrow
Not having a mosaic I am only speaking from general knowledge of hobby
printers but calibration is a product of material and machine. For a
given print material and printer, if you assemble it you WILL be doing
calibration prints, they are fairly painless given a good starting
profile and a well designed printer. As far as further calibration,
that may be necessary should you change print material, radically
change the printing environment or decide you are printing to a
different goal, (aesthetics, speed, structural integrity) or decide
you want to focus on some areas printers have problems with like
overhangs, support structure or bridges. It is always possible to
recalibrate your printer and the need to do so and difficulty involved
will only get easier as software improves. That's my two bits as far
as calibration is concerned.
printers but calibration is a product of material and machine. For a
given print material and printer, if you assemble it you WILL be doing
calibration prints, they are fairly painless given a good starting
profile and a well designed printer. As far as further calibration,
that may be necessary should you change print material, radically
change the printing environment or decide you are printing to a
different goal, (aesthetics, speed, structural integrity) or decide
you want to focus on some areas printers have problems with like
overhangs, support structure or bridges. It is always possible to
recalibrate your printer and the need to do so and difficulty involved
will only get easier as software improves. That's my two bits as far
as calibration is concerned.
Fred
Aug 21, 2011, 5:32:04 PM8/21/11
to MakerGear - Make Today, Change Tomorrow
Luke, can you speak to calibration? How do you do it?
I have printed the 20mm calibration cubes, they come out great,
but I still believe I have some issues with flow.
What do you do to calibrate?
Thanks
Fred
> > Thanks!- Hide quoted text -
>
> - Show quoted text -
I have printed the 20mm calibration cubes, they come out great,
but I still believe I have some issues with flow.
What do you do to calibrate?
Thanks
Fred
>
> - Show quoted text -
Luis E. Rodriguez
Aug 22, 2011, 5:22:03 PM8/22/11
to make...@googlegroups.com
I would refer to Dave Durrant's blog as he is GOD go Skeinforge. Feel free to contact him, he is super duper nice. Except when he is a butt. That was bait for him. :)
Luke
Aug 22, 2011, 6:17:17 PM8/22/11
to MakerGear - Make Today, Change Tomorrow
Hi Fred,
Getting the proper flow rate can be very tricky, it is also hard to
judge as many other factors masquerade as flow rate problems. But here
is my procedure, of which all steps may or may not apply to you. I use
a stepper extruder and for a given layer thickness, width / thickness,
and feed-rate I only adjust flow-rate.
1. Verify extruder ability to reliably and consistently drive
filament.
When working with a new filament driver, hot-end or material it is
important to ensure your filament moves through the extruder reliably
and at the same rate. This means finding the right combination of flow
rate, temperature and pressure against the filament. I adjust these
until I know the maximum flow-rate I can run my extruder at before I
skip steps, strip filament, or experience inconsistent flow. Believe
it or not it IS possible for your filament to slow down and speed up
as it encounters resistance in the hot-end. I have a bearing pushing
my filament against the driver and compare monitor its rotation. If
the bearing spins at a constant rate, I'm good.
2. Level your platform.
I print http://www.thingiverse.com/thing:10402 and measure the
perimeter heights with calipers. It may be helpful to stop your print
before infill to do this. More times than I'd like to admit, a problem
I had been chasing down somewhere else was solved by a good level
build surface.
3. Find a happy starting Z height and first layer feedrate.
You can't calibrate your printer if your plastic won't stick. To start
I peg my first layer to print at half my layer height and my feed-rate
at half operating feed-rate. If your sticking great and can take your
nozzle higher it will help later in the calibration process to not
have all the extra plastic gooshing up from being so low. In general
if things aren't sticking, go slower and lower, in that order.
4. Print a simple object.
Find an object without reversals, overhangs or bridges; the 20mm cube
is fine as long as you aren't going too fast or too hot as you will
run into other problems that make accurate flow analysis difficult. (I
know the tried and true way is to hit full infill and see how it goes
but I stopped doing that as it doesn't tell the whole story. Either
way, you need to watch your infill pattern as one of them even at full
infill gives a semi-solid layer periodically depending on your repeat
layer settings. I believe its line, so if your using hex or another,
full infill should actually give you 100% infill.)
For me, the most important things are perimeter trace width and solid
layer surface finish. These together give me the best looking and
accurate parts. I set infill to 30%, solid layers to 3 and start
printing, extra shells to 1 and hit print. Half way through the print,
when the nozzle is half way around the perimeter the second time I
abort the print and raise the nozzle. Let it cool a bit so you don't
distort it when removing it and to let any contraction occur then pry
it off and grab your calipers. Measure the width of your extra loop
and the combined width of your extra loop and perimeter, you should
have both available if you stopped your print at the right time. They
should be the same percentage off from your expected width. Say, I
measure .55mm on an expected width of .5mm I adjust my flow rate by .
5 / .55 X Current Flow-rate = New Flow-rate. This doesn't work
perfectly because trace geometry isn't orthogonal but it gets you
close enough to focus on getting your top layer looking nice.
Start a new print and either measure your trace widths as before to
see how things changed or let the print finish. I've found it takes AT
LEAST 3 solid surface layers to have created enough support so that
the top layer accurately reflects your flow rate. Take your print and
look at the top, if you can see between the lines, increase your flow
rate, if the lines look a little crowded, ease up. If you see gaps
only at the beginning and tale ends of lines you can increase Infill
Perimeter Overlap but only to a certain degree before you run into
other troubles. If I've nailed my perimeter trace-widths I find I can
still adjust my flow-rates to make the solid layers look nice without
changing my perimeter width, it's a handy by product of the fact that
those traces aren't perfect square cross-sections and can accept a
little more or a little less plastic without changing their width
substantially. Its possible to get very smooth side and top surfaces
by following this procedure.
5. Track your other settings.
Namely reversal, width over thickness and loop order.
Reversal make it look like you have terrible flow settings when you
really don't. To aggressive and your extruder can skip a step stopping
all flow for a second or two...its horrible. Lots of small loops like
screw holes in a print are hazardous if you have long pull and push
times as they are activated BEFORE the extruder starts and finishes.
If the hole is too small or you run too fast your extruder stops
before its finished the circle and your ship is sunk.
Width over Thickness is a very important setting to track. After die
swell, I measure the diameter of extruded material from my .4mm nozzle
to be .44mm. I bump that number up to .5mm and use that for my trace
widths. This gives me even division of metric units and with a layer
thickness of .2mm allows for tremendous overhang. You have to remember
that your ability to overhang is directly related to Width over
Thickness. With a W/T of 2.5 (.5mm Width / .2mm Height) an overhang of
45 degrees still gives me 60% of a trace width laying on the trace
below it.
In general, I get better looking prints by setting loop order to
Perimeter, Loops, Infill though it reduces my ability to overhang when
compared to Loop, Perimeter, Infill. I also get better looking prints
by printing hollow with zero infill, in THAT case, your real overhangs
can come from the internal cavity and convex prints such as a simple
sphere will overhang better with Perimeter, Loop, Infill as the
overhang is internal.
Anyways, I hope that answers some questions, that last bit isn't so
much related to flow-rate calibration but I do find myself in those
settings often when doing specialized prints. For instance, our
company has been printing a large volume of Plastic T-slot beams, of
which getting the strongest, straightest and fastest prints involved
massively changing the standard so that each layer is a series of
loops bypassing infill completely. At one point we turned clip up to 7
so that the very small amount of extra plastic extruded when changing
to a new loop didn't accumulate uncontrollably at the end of 600
layers.
If you have any other questions please ask,
Luke
Getting the proper flow rate can be very tricky, it is also hard to
judge as many other factors masquerade as flow rate problems. But here
is my procedure, of which all steps may or may not apply to you. I use
a stepper extruder and for a given layer thickness, width / thickness,
and feed-rate I only adjust flow-rate.
1. Verify extruder ability to reliably and consistently drive
filament.
When working with a new filament driver, hot-end or material it is
important to ensure your filament moves through the extruder reliably
and at the same rate. This means finding the right combination of flow
rate, temperature and pressure against the filament. I adjust these
until I know the maximum flow-rate I can run my extruder at before I
skip steps, strip filament, or experience inconsistent flow. Believe
it or not it IS possible for your filament to slow down and speed up
as it encounters resistance in the hot-end. I have a bearing pushing
my filament against the driver and compare monitor its rotation. If
the bearing spins at a constant rate, I'm good.
2. Level your platform.
I print http://www.thingiverse.com/thing:10402 and measure the
perimeter heights with calipers. It may be helpful to stop your print
before infill to do this. More times than I'd like to admit, a problem
I had been chasing down somewhere else was solved by a good level
build surface.
3. Find a happy starting Z height and first layer feedrate.
You can't calibrate your printer if your plastic won't stick. To start
I peg my first layer to print at half my layer height and my feed-rate
at half operating feed-rate. If your sticking great and can take your
nozzle higher it will help later in the calibration process to not
have all the extra plastic gooshing up from being so low. In general
if things aren't sticking, go slower and lower, in that order.
4. Print a simple object.
Find an object without reversals, overhangs or bridges; the 20mm cube
is fine as long as you aren't going too fast or too hot as you will
run into other problems that make accurate flow analysis difficult. (I
know the tried and true way is to hit full infill and see how it goes
but I stopped doing that as it doesn't tell the whole story. Either
way, you need to watch your infill pattern as one of them even at full
infill gives a semi-solid layer periodically depending on your repeat
layer settings. I believe its line, so if your using hex or another,
full infill should actually give you 100% infill.)
For me, the most important things are perimeter trace width and solid
layer surface finish. These together give me the best looking and
accurate parts. I set infill to 30%, solid layers to 3 and start
printing, extra shells to 1 and hit print. Half way through the print,
when the nozzle is half way around the perimeter the second time I
abort the print and raise the nozzle. Let it cool a bit so you don't
distort it when removing it and to let any contraction occur then pry
it off and grab your calipers. Measure the width of your extra loop
and the combined width of your extra loop and perimeter, you should
have both available if you stopped your print at the right time. They
should be the same percentage off from your expected width. Say, I
measure .55mm on an expected width of .5mm I adjust my flow rate by .
5 / .55 X Current Flow-rate = New Flow-rate. This doesn't work
perfectly because trace geometry isn't orthogonal but it gets you
close enough to focus on getting your top layer looking nice.
Start a new print and either measure your trace widths as before to
see how things changed or let the print finish. I've found it takes AT
LEAST 3 solid surface layers to have created enough support so that
the top layer accurately reflects your flow rate. Take your print and
look at the top, if you can see between the lines, increase your flow
rate, if the lines look a little crowded, ease up. If you see gaps
only at the beginning and tale ends of lines you can increase Infill
Perimeter Overlap but only to a certain degree before you run into
other troubles. If I've nailed my perimeter trace-widths I find I can
still adjust my flow-rates to make the solid layers look nice without
changing my perimeter width, it's a handy by product of the fact that
those traces aren't perfect square cross-sections and can accept a
little more or a little less plastic without changing their width
substantially. Its possible to get very smooth side and top surfaces
by following this procedure.
5. Track your other settings.
Namely reversal, width over thickness and loop order.
Reversal make it look like you have terrible flow settings when you
really don't. To aggressive and your extruder can skip a step stopping
all flow for a second or two...its horrible. Lots of small loops like
screw holes in a print are hazardous if you have long pull and push
times as they are activated BEFORE the extruder starts and finishes.
If the hole is too small or you run too fast your extruder stops
before its finished the circle and your ship is sunk.
Width over Thickness is a very important setting to track. After die
swell, I measure the diameter of extruded material from my .4mm nozzle
to be .44mm. I bump that number up to .5mm and use that for my trace
widths. This gives me even division of metric units and with a layer
thickness of .2mm allows for tremendous overhang. You have to remember
that your ability to overhang is directly related to Width over
Thickness. With a W/T of 2.5 (.5mm Width / .2mm Height) an overhang of
45 degrees still gives me 60% of a trace width laying on the trace
below it.
In general, I get better looking prints by setting loop order to
Perimeter, Loops, Infill though it reduces my ability to overhang when
compared to Loop, Perimeter, Infill. I also get better looking prints
by printing hollow with zero infill, in THAT case, your real overhangs
can come from the internal cavity and convex prints such as a simple
sphere will overhang better with Perimeter, Loop, Infill as the
overhang is internal.
Anyways, I hope that answers some questions, that last bit isn't so
much related to flow-rate calibration but I do find myself in those
settings often when doing specialized prints. For instance, our
company has been printing a large volume of Plastic T-slot beams, of
which getting the strongest, straightest and fastest prints involved
massively changing the standard so that each layer is a series of
loops bypassing infill completely. At one point we turned clip up to 7
so that the very small amount of extra plastic extruded when changing
to a new loop didn't accumulate uncontrollably at the end of 600
layers.
If you have any other questions please ask,
Luke
Ahmet TURAN
Aug 23, 2011, 8:30:32 AM8/23/11
to make...@googlegroups.com
Why dont you try SFACT?
That has very easy calibration…
ddurant
Aug 23, 2011, 10:17:15 AM8/23/11
to make...@googlegroups.com
> Except when he is a butt.
:P
I'd be pretty surprised if Mosaic didn't support volumetric 5D, which takes away a lot of the mystery of skeinforge. Measure you right filament diameter, pick whatever feed/WoT/layer-height values you want and it just does the right thing..
Mike Payson
Aug 23, 2011, 2:31:08 PM8/23/11
to make...@googlegroups.com
Nothing supports 5D anymore. 5D was not a real thing, it was a badly implemented hack where "acceleration" was treated as an axis. It was only ever supported by RepRapHost and the RepRap 5D firmware. All other firmwares are 4D (XYZ and E-- though some support more than one E, so technically could be viewed as nD). Some 4D firmwares support acceleration and others don't, but they are still all 4D.
Volumetric has little to do with the Firmware. It is related to how the model is skeined, not how the firmware works. Because you need to be able to precisely control the extrude speed it does require a stepper extruder, but Gen3 with a stepper extruder was perfectly capable of doing Volumetric.
Rob Giseburt
Aug 23, 2011, 6:16:52 PM8/23/11
to make...@googlegroups.com
"5D" is one of those things where the nomenclature stuck, even after the thing being referred to changed enough to make it inaccurate.
Generally, when people refer to 5D anymore, they really do just mean X, Y, Z, and E. Yes, the original 5 Ds included acceleration, but that is rarely used any more. At least, not in that simplistic form.
"Volumetric 5D" refers to using E as an axis instead of RPM values, and calculating the E based on volume of filament needed. I know because I coined the term.
Some firmwares don't support E values, or "5D." Others don't support RPM values. We are currently divided.
-Rob
ddurant
Aug 23, 2011, 8:15:51 PM8/23/11
to make...@googlegroups.com
Yeah, what Rob said!
And as for volumetric, non-"volumetric 5D" printing, I think I've heard of that stuff too.
Rob Giseburt
Aug 23, 2011, 9:16:40 PM8/23/11
to make...@googlegroups.com
Yes, Dave and I worked both simultaneously and occasionally together
on the Volumetric part. It's still, as far as I'm concerned, not done.
on the Volumetric part. It's still, as far as I'm concerned, not done.
Since the Mosaic uses RAMPS, then the "standard" RepRap firmware,
along with several others, will do "5D" without modification.
-Rob
Luke
Aug 24, 2011, 1:55:13 AM8/24/11
to MakerGear - Make Today, Change Tomorrow
Well by 5D or other method, your still tweaking numbers in the end to
give you results. 5D perhaps gives you less numbers to change for a
given iteration but to calibrate good fudge factors you will
ultimately be adjusting SOMETHING. For me I find the best thing to get
at is your flow rate, however that be tackled in skeinforge.
On Aug 23, 6:16 pm, Rob Giseburt <giseb...@gmail.com> wrote:
> Yes, Dave and I worked both simultaneously and occasionally together
> on the Volumetric part. It's still, as far as I'm concerned, not done.
>
> Since the Mosaic uses RAMPS, then the "standard" RepRap firmware,
> along with several others, will do "5D" without modification.
>
> -Rob
>
give you results. 5D perhaps gives you less numbers to change for a
given iteration but to calibrate good fudge factors you will
ultimately be adjusting SOMETHING. For me I find the best thing to get
at is your flow rate, however that be tackled in skeinforge.
On Aug 23, 6:16 pm, Rob Giseburt <giseb...@gmail.com> wrote:
> Yes, Dave and I worked both simultaneously and occasionally together
> on the Volumetric part. It's still, as far as I'm concerned, not done.
>
> Since the Mosaic uses RAMPS, then the "standard" RepRap firmware,
> along with several others, will do "5D" without modification.
>
> -Rob
>
Mike Payson
Aug 24, 2011, 3:22:29 AM8/24/11
to make...@googlegroups.com
Ok, so 5D is not really 5D, but you call it that for historical reasons, and volumetric 5D is, in effect, using E steps to calculate the volume of plastic extruded... Ok, fair enough.
But why call it volumetric 5d? Why not just call it volumetric and be done with it, and put the antiquated, inaccurate and confusing term to rest? I suppose that other volume measurement systems are possible, but to the best of my knowledge, no other volumetric systems exist at present, so you are not really introducing any ambiguity by dropping the term, and in fact you are clearing up a consistent source of confusion since the RepRap 5D firmware was very different in every possible way except that it also used e-steps to calculate volume.
FWIW, I also disagree that the term "5d" stuck. It is basically never used in the IRC community. There are modern FW's that support acceleration and volumetric extrusion (basically all the recent ones, but mainly Sprinter, Teacup and SJFW), older ones that are largely obsolete but that are still volumetric (Mainly Tonokip, but almost no one is using these anymore), and all the other obsolete stuff that no one cares about (5d, Makerbot FW's, etc). Since virtually everyone is using a volumetric FW at this point, there is little point in the distinction. Virtually the only time the term even comes up is when we explain to people why they need to upgrade to SF40+.
Owen M Collins
Aug 24, 2011, 9:13:42 AM8/24/11
to make...@googlegroups.com
So as someone setting up their Makergear Stepper Plastruder on a MB Cupcake, are you saying that I won't get the full potential using the MB firmwares, or those Firmwares that support volumetric Extrusion in conjunction with the newest releases of SF make the best print quality?
I have found this discussion really interesting by the way, thanks.
O.
Joel Chia
Aug 24, 2011, 9:30:44 AM8/24/11
to make...@googlegroups.com
It probably means that it's trickier to get your profile exactly right. And changing to change a value means that you have to change 3 or 4 more values elsewhere to accommodate the first change.
I'm not sure what electronics you're using...
As for firmware, I asked this question in the MB google groups re Gen4 and found out that the MB firmware has volumetric support... just need to change your machine settings.
If you're using Gen3, Teacup supposedly works on Gen3
-Joel
Owen M Collins
Aug 24, 2011, 9:52:46 AM8/24/11
to make...@googlegroups.com
I'm still on Gen3. Does the choice of Firmware necessitate a change in RepG? Or would I have to use something different?
Thanks,
O.
Triffid Hunter
Aug 24, 2011, 9:54:23 AM8/24/11
to make...@googlegroups.com
In theory, ReplicatorG should work. in practice, it has some nasty
bugs that make it rather unusable with modern firmwares. We're all
using pronterface ( http://github.com/kliment/Printrun ) these days.
bugs that make it rather unusable with modern firmwares. We're all
using pronterface ( http://github.com/kliment/Printrun ) these days.
Owen M Collins
Aug 24, 2011, 10:04:40 AM8/24/11
to make...@googlegroups.com
Is that cross platform? I am using a mac. HOw different is it practice?
Sometimes it's hard to keep up with everything, I guess that is the price of living in the future...
I'll look into that thanks,
O.
ddurant
Aug 24, 2011, 10:18:09 AM8/24/11
to make...@googlegroups.com
> Yes, Dave and I worked both simultaneously and occasionally together on the Volumetric part.
Yeah but you're the one that got it into skeinforge (or at least got Enrique to get it in). The work I did was just because I was sick of making those damned 20mm cubes.
People using SF40+ and the Dimension module should thank Rob for making their configuration lives sooooo much easier, if they haven't already..
> It's still, as far as I'm concerned, not done
Could be but it's still a lot easier than it used to be. :)
Triffid Hunter
Aug 24, 2011, 11:03:58 AM8/24/11
to make...@googlegroups.com
On Thu, Aug 25, 2011 at 12:04 AM, Owen M Collins
<ccstud...@gmail.com> wrote:
> Is that cross platform? I am using a mac.
<ccstud...@gmail.com> wrote:
> Is that cross platform? I am using a mac.
yes, it's written in python. There's a readme somewhere that lists
dependencies, basically wxpython, pyserial, readline iirc.
join us in irc://irc.freenode.net/#reprap , I'm sure there are quite a
few running pronterface on macs.
Luis E. Rodriguez
Aug 24, 2011, 12:18:58 PM8/24/11
to make...@googlegroups.com
Thank you Rob! And I still make cubes!
I definitely alter settings a lot less in SF40. I update filament diameter as best as I can over an average of different places in the filament and then things like fill and shells based on the object.
I am waiting patiently for acceleration in the next MB firmware as I like what I see coming from the Ultimaker.
I'm wondering if it means we need to use LM8UU linear bearings or similar instead though? Shaking your bot around like that starts to loosen screws!
Luis E. Rodriguez
Luis E. Rodriguez
Mike Payson
Aug 24, 2011, 1:41:42 PM8/24/11
to make...@googlegroups.com
On Wed, Aug 24, 2011 at 6:13 AM, Owen M Collins <ccstud...@gmail.com> wrote:
So as someone setting up their Makergear Stepper Plastruder on a MB Cupcake, are you saying that I won't get the full potential using the MB firmwares, or those Firmwares that support volumetric Extrusion in conjunction with the newest releases of SF make the best print quality?I have found this discussion really interesting by the way, thanks.
Yes, that is correct. Going to a stepper extruder alone gives you much more consistent results, but to really see the benefits you need to be running firmware and slicing software that are designed to take advantage of it. Like Joel said, for Gen3, I believe your only option is Teacup.
Reply all
Reply to author
Forward
0 new messages