Can't get thin walled objects to print correctly, Help Needed

[DiyAddicts]

I am proud owner of ToM 101 and I am getting things going again after receiving a replacement extruder motor. Anyways, I am getting very good print quality on most objects. I am though having a nightmare with objects that have thinner walls. Both my perimeter width over thickness and infill width over thickness 1.45. Also have my perimeter infill overlap 0.1. These were the only suggestions I could find. I attached some pictures of a few dodecahedrons and an arm to the lovable android thing #1902. They seem to do ok for the first few layers then start to drop some layers too far outside the previous layer. Also happened on the factual tree thing #4340. Another thing I noticed on the tree and android arm is that when putting down the layers it seems like it is not coming up far enough and dragging through the whole object just a little causing it to wobble while it is printing. Any help would be greatly appreciated. I hope everyone had a great New Years!

AM

[DiyAddicts]

I'll give that a shot here in about an hour when I get home. It just seems odd that all the other prints I do turn out very well.

[ddurant]

What temperature are you running at? Looks to me like it might be a
little high.. It also looks like the prints start ok then go sorta
pear shaped. It could be that you're warming up to a good temperature
but have skeinforge set to a higher one, which messes things up as the
print progresses and the temp reaches the set value.

Also, your pix are all of the sides of objects but the term 'wall
thickness' sorta only applies to views from the top.

> it seems like it is not coming up far enough and dragging through the whole object

That's usually an indication that you need to turn the feed rate up.
Too much plastic = turn feed rate up a little; too little plastic =
turn feed rate down a little. I'd get the temperature sorted first
though, because the prints do start out ok then go south.

> I hope everyone had a great New Years!

Back at ya!

[The Ruttmeister]

I'd check the various temp settings, but I'd also check the belt
tension..

[Charles Edward Pax]

Which version of Skeinforge are you using?

Charles Edward Pax
 blog: http://charlespax.com/
 twitter: http://twitter.com/charlespax

> --
> 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.
>

[DiyAddicts]

>On Sunday, January 2, 2011 8:36:43 PM UTC-6, Charles Edward Pax wrote:

Which version of Skeinforge are you using?

I am using Skeingorge 35 that is packaged in RepG 23

[DiyAddicts]

Sorry for the delayed reply, I'm not getting my email notifications for some reason.

>What temperature are you running at?

My temps are set at the default 225 still.

I spent about four hours last night printing test android arms. I have gotten a lot closer I think, but I am honestly a little stumped on what got me closer and what to keep tweaking. After about 15 prints I got the attached pictures with the following settings:

Feed rate: 30
w/t:            1.57
layer thickness: .38

It now looks like they are starting off very bad and then finishing up ok. It does look like the feed rate might be a little high by looking at the tops.

>It could be that you're warming up to a good temperature
but have skeinforge set to a higher one, which messes things up as the
print progresses and the temp reaches the set value.

Forgive me for my ignorance but is there anyway to monitor what the temperature is during the print? All I can test right now is extruding manually and the temp seems to say +/-2-3 degrees

I appreciate the help everyone.

[ddurant]

> Forgive me for my ignorance but is there anyway to monitor what the
> temperature is during the print?

Not really.. There used to be an option in repg to display the current
temperature but it caused problems so they removed it. Hopefully, it
will find its way back in someday - I miss the option (but not the
problems!)

Use a text editor to open up one of the gcode files you're printing
and search for "M104" (set extruder temperture) and "M109" (set HBP
temperature) commands. They'll look something like "M104 S210 T0",
which sets the extruder to 210C.

Do the temperatures settings in the gcode look correct to you? What
sort of values do you see in there?

> Feed rate: 30

The problem could also be that you're printing too a little too slowly
- low feed rates can have more trouble with overhangs than higher
ones. 30 isn't overly slow but it's certainly not fast.

Have you tried printing something that's thin-walled but does not have
overhangs? http://www.thingiverse.com/thing:3057 might be a good test
object.

>  photo 1[1].JPG
> 200KViewDownload
>
>  photo 2[1].JPG
> 170KViewDownload

[DiyAddicts]

>Use a text editor to open up one of the gcode files you're printing
and search for "M104" (set extruder temperture) and "M109" (set HBP
temperature) commands. They'll look something like "M104 S210 T0",
which sets the extruder to 210C.

Do the temperatures settings in the gcode look correct to you? What
sort of values do you see in there?

M104 is at 225 and M109 is at 125

>The problem could also be that you're printing too a little too slowly
- low feed rates can have more trouble with overhangs than higher
ones. 30 isn't overly slow but it's certainly not fast.

I had been printing at 22 mm/sec and getting pretty good results. (I got that number from measuring the extrusion as posted on the makerbot blog I believe it was). But I am up to trying anything :)

>Have you tried printing something that's thin-walled but does not have
overhangs?  http://www.thingiverse.com/thing:3057 might be a good test
object.

I just printed it and am printing one right now at my original settings. I will upload the settings shortly.

[Aaron Double]

I'm curious how the wobble arrester comes out. Those are a pain:)

On Jan 3, 8:32 pm, DiyAddicts <adam_mar...@att.net> wrote:
> >Use a text editor to open up one of the gcode files you're printing
>
> and search for "M104" (set extruder temperture) and "M109" (set HBP
> temperature) commands. They'll look something like "M104 S210 T0",
> which sets the extruder to 210C.
>
> Do the temperatures settings in the gcode look correct to you? What
> sort of values do you see in there?
>
> M104 is at 225 and M109 is at 125
>
> >The problem could also be that you're printing too a little too slowly
>
> - low feed rates can have more trouble with overhangs than higher
> ones. 30 isn't overly slow but it's certainly not fast.
>
> I had been printing at 22 mm/sec and getting pretty good results. (I got
> that number from measuring the extrusion as posted on the makerbot blog I
> believe it was). But I am up to trying anything :)
>
> >Have you tried printing something that's thin-walled but does not have
>

> overhangs?  http://www.thingiverse.com/thing:3057might be a good test

[DiyAddicts]

Holy Balls this is tedious! :) Tweaking Makerbot and untreated ADD do not mix well. Anyways I have printed 5 Z wobble arresters (and one printing now) playing with some settings. I attached the photos with the settings I used written on paper in the picture. And sorry they uploaded in the wrong order. There are some with way too little plastic and others with too much in places and not enough in others. I don't even know if I am going in the right direction yet (good thing I have 11 lbs of ABS left!) Also one other thing I notice with this print is when it is traveling it does so right over the previous layer and seems to rub on the top of the previous layer, weird when it is not even extruding. Thanks again for all the help guys, appreciate your patience.

AM

[Danial]

I can't say for sure, but your prints look a fair bit like mine did when I had the Z set way too high off the build platform. I had mine a full .5 too high. Try making sure you are far enough down. Run the ToM calibration gcode provided in RepG23. It is how I learned I was way too high.

[DiyAddicts]

I've done that a few times and it seems like I have it pretty close. The raft lays down without a problem. Did your raft lay down fine even when you had the Z max too high?

AM

[ddurant]

> I attached the photos

Is it just the way ABS takes pictures or does the right side of each
one of those sorta curve in on itself? If they are curving in, your
temperature is too high - try dropping it to 215 and seeing how that
works.

I think you're actually not too far off but your thread width is too
big to print that part correctly - sorry, I'd forgotten just how thin
those walls are. As Aaron said, that's a tricky Thing to print. That
used to be my thin-wall test piece but it's been a while since I've
printed one..

It looks like you're poking at width over thickness. Make sure you do
both perimeter w/t in Carve and infill w/t in Fill, unless you know
you want different values.

>  photo 1.JPG
> 189KViewDownload
>
>  photo 2.JPG
> 179KViewDownload
>
>  photo 3.JPG
> 179KViewDownload
>
>  photo 4.JPG
> 202KViewDownload
>
>  photo 5.JPG
> 175KViewDownload

[Danial]

Well, I thought they did, they looked pretty, but they refused to pull off correctly. The thick bottom layer would, but the second thin layer always stuck to the object. Now that I have my Z max lower, the raft is solid on the bottom, no gaps between each thick pass, and the two thin layers on top of it are largely inset. That said, now the raft refuses to pull off the object at all, so that's not perfect either.

[DiyAddicts]

>Is it just the way ABS takes pictures or does the right side of each
one of those sorta curve in on itself? If they are curving in, your
temperature is too high - try dropping it to 215 and seeing how that
works.

It is more like they curve outward just a bit. The camera did make it seem that way though.

>I think you're actually not too far off but your thread width is too
big to print that part correctly - sorry, I'd forgotten just how thin
those walls are. As Aaron said, that's a tricky Thing to print. That
used to be my thin-wall test piece but it's been a while since I've
printed one..

When you say my thread width is too big do you mean that is a setting I need to change or the walls are too thin for my nozzle? Is there something else you would recommend to print? What makes this so hard is I originally calibrated with the 20mm cube and thought I had my settings dialed in and then when I move to thin walled/overhangs it is horrible.

>It looks like you're poking at width over thickness. Make sure you do
both perimeter w/t in Carve and infill w/t in Fill, unless you know
you want different values.

I messed with w/t a little and always kept both the same. I'm trying to only adjust one setting at a time to keep the variables down. It is quiet the process.

AM

[ddurant]

> It is more like they curve outward just a bit.

Either way, a straight vertical side printing as a curve usually means
temperature is too high..

> When you say my thread width is too big...

I had a really long winded answer going then changed my mind..

Can you do one more try at that object with the 3 "Extra Shells"
settings in Fill set to 0?

[DiyAddicts]

>Can you do one more try at that object with the 3 "Extra Shells"
settings in Fill set to 0?

Hey I will try anything (almost) :) I printed it out with those set to 0 and it printed the whole object at an angle. attached photo of it. The other photo is of a print I did late last night that turned out the best so far, but is still way off.

[ddurant]

> the best so far..

Notice the bits on the bottom of that one - the arms that end in the
bolt holes. Those arms should be solid but the picture shows a gap in
the wall.

Turning off the extra shells leaves skeinforge enough room to fill in
that gap. Even though the shells=0 version is a bit of a leaning
tower, I think it's still the "more correct" print. (In general, btw,
I leave the extra shell settings where they are and only turn them
down for Things that have very thin walls.)

A print leaning like that usually means you have hardware slipping. I
really don't know much about ToMs but the most common causes for that
are belts being too loose or the nozzle catching on plastic blobs.
Could also be that the pot on that stepper driver is turned down too
low. The print doesn't look very blobby to me so I'd check that you
don't have any loose belts then maybe turn that one stepper up a
little bit.

Did you turn the temperature down a bit? Did that help or make things
worse?

>  photo 2.JPG
> 205KViewDownload
>
>  photo 1.JPG
> 172KViewDownload

[Zip Zap]

Anyone have any suggestions on how to print out DXF files on paper?  I've got Turbo Cad but having problems getting anything to print out.

[DiyAddicts]

>Did you turn the temperature down a bit? Did that help or make things
worse?

I haven't had a chance to do this yet, I will tonight though.

>A print leaning like that usually means you have hardware slipping. I
really don't know much about ToMs but the most common causes for that
are belts being too loose or the nozzle catching on plastic blobs.
Could also be that the pot on that stepper driver is turned down too
low. The print doesn't look very blobby to me so I'd check that you
don't have any loose belts then maybe turn that one stepper up a
little bit.

The belts and all seem ok. That was the first print I had lean like that and it was caused from the change to the shells. I would think if I had a belt and/or voltage problem that it would affect all my prints in some way. I have printed out about 15 other objects that turned out well. Please correct me if I am wrong on this though because I am still learning. Thanks again!

AM

[DiyAddicts]

I think you should maybe post a new question on this. Not really relevant to the thread.

Regards,

AM

[Achilles Boiser]

Anyone have any suggestions on how to print out DXF files on paper?  I've got Turbo Cad but having problems getting anything to print out.  I'm trying to print out the DXF files for the ThingOMatic.

[ddurant]

The leaning tower thing means that an axis is slipping. If you're
pretty sure that's not caused by a mechanical issue like loose belts
or the nozzle catching on a blog, I'd turn up the pot on the stepper
driver. That will give it a little more push, which you may need for a
print like this.

> ...was caused from the change to the shells

Yes but not exactly. Fewer shells meant it was able to do the infill
and the infill on that object is pretty thin. Lots & lots of little
zig-zags, which may be too much for the pot setting you're at now.

[Joheinz]

Try Qcad, it has good DXF support.
Markus

2011/1/5 Achilles Boiser <achille...@gmail.com>:

> Anyone have any suggestions on how to print out DXF files on paper?  I've
> got Turbo Cad but having problems getting anything to print out.  I'm trying
> to print out the DXF files for the ThingOMatic.
>

> --
> 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.
>

--
http://justjoheinz.posterous.com/
http://softwarepoets.org
http://twitter.com/joheinz

[Aaron Double]

Why are people hijacking this discussion about getting prints to print
correctly to talk about dxf files?

Can't you make your own post about that subject?

On Jan 5, 12:25 am, Joheinz <justjohe...@googlemail.com> wrote:
> Try Qcad, it has good DXF support.
> Markus
>

> 2011/1/5 Achilles Boiser <achillesboi...@gmail.com>:

[Z LeHericy]

odd... I'm seeing them as separate posts... but it does seem that several threads have been accidentally woven together...

-Zeno LeHericy

//((=:Z:=))\\
INVENTIONS
Technologies
zinventions.com

[DiyAddicts]

>The leaning tower thing means that an axis is slipping. If you're
pretty sure that's not caused by a mechanical issue like loose belts
or the nozzle catching on a blog, I'd turn up the pot on the stepper
driver. That will give it a little more push, which you may need for a
print like this.

I spent a ton of time on this and I think have ruled it out. But after digging into this for many hours I am now thinking it is mostly a temperature based problem after looking into it a little closer. I never noticed how much variation was there before, but I have some pretty big swings in temperature on the extruder:

When set to 225C and extruding at full PWM it would constantly swing from 217C to 235C

When set to 216C it would swing from 208C to 218C.

I don't know what the acceptable temperature variations are, but these seem like a lot to me. I would like some feedback, but I'm guessing it is time to start chiming away at the PID settings.

[ddurant]

> When set to 225C and extruding at full PWM it would constantly swing from 217C to 235C

Sounds like your PID values needed some lovin'. This is something I
know little about - others here understand it, though.

Unless the temperature is causing blobs (unlikely) and the nozzle is
catching on them, that axis slip really probably is a pot that's
turned down too low.

[DiyAddicts]

>Sounds like your PID values needed some lovin'. This is something I
know little about - others here understand it, though.

Yeah looks like I have a project for this evening :)

>Unless the temperature is causing blobs (unlikely) and the nozzle is
catching on them, that axis slip really probably is a pot that's
turned down too low.

I think I have ruled out an axis slip. I even ended up turning the pot all the way up (got a little warm)  just to make sure. There is no excess friction and belts have really good tension. The only time I can reproduce the lean with several different models is by turning down the extra shells. So I'm going to hope that getting the temperature closer will help get me closer.

Does anybody know how much variance is ok on the extruder temperature?

[DiyAddicts]

>Sounds like your PID values needed some lovin'. This is something I
know little about - others here understand it, though.

Yeah looks like I have a project for this evening :)

>Unless the temperature is causing blobs (unlikely) and the nozzle is
catching on them, that axis slip really probably is a pot that's
turned down too low.

[ddurant]

> I think I have ruled out an axis slip

You ruled out it being caused by a pot that's too low!! The axis *IS*
slipping - you even posted a picture of it.

Has anybody else with a ToM tried to get a good print of that wobble
arrestor?

[Andrew Plumb]

Make sure you have them set to the ToM-specific defaults listed here: http://wiki.makerbot.com/thingomatic-doc:calibration

...not the relay-driven Mk5 defaults listed elsewhere.

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.

--

Makerbot Number Nine... #9... 0x09... 0o11... 0b1001... 
http://clothbot.com/wiki/MakerBotNumberNine

[DiyAddicts]

>You ruled out it being caused by a pot that's too low!! The axis *IS*
slipping - you even posted a picture of it.

If I had an axis slipping it should slip no matter what the shell settings are correct? I printed several different objects last night and they would show no sign of slipping/leaning. Then I would print the same objects with the extra shells at zero and they would all lean. I change ONE variable and objects lean, I change it back and they don't. In my mind that seems pretty straight forward, although I am no expert at this.

[DiyAddicts]

>Make sure you have them set to the ToM-specific defaults listed here: http://wiki.makerbot.com/thingomatic-doc:calibration

...not the relay-driven Mk5 defaults listed elsewhere.

I had seen that page, but missed those values for some reason. Mine are pretty far off from those so I will try them when I get home. Thanks!

AM

[Aaron Double]

It'll only slip when the stage has to go rapidly back and forth,
right. Do you know if it is slipping while printing? or when
traveling? If when printing, I would try slowing your feedrate and
travelrate, if just when travelling, just slow down travel rate.

Something else that might be interesting would be to run it with less
shells and comb turned off. You'll should get a bunch of strings but
you'll have much less travel. (that part will print much faster also)

That part is a pain.

To see an example of the rapid back and forth take a peek at this
video:

http://www.flickr.com/photos/48982649@N07/5324656172/

[ddurant]

> http://www.flickr.com/photos/48982649@N07/5324656172/

Next thingiverse contest: bobble heads to mount on the top of our
bots!

(don't take offense - my machine shakes waaaay more than that)

> > AM- Hide quoted text -
>
> - Show quoted text -

[Aaron Double]

It's funny, once I put in the larger platform, the shaking greatly
increased, but then it's a chunk of aluminum 110x130x6mm. Had to turn
up the pots on the stepper boards as well. With Rob's PSMD it's at
least quietly shaking...

The bobble head would be funny but I think it would get old in like 5
minutes.

[ddurant]

> If I had an axis slipping it should slip no matter what the shell settings
> are correct?

Nope. Sorta like Aaron said above, it's the rapid back & forth
movement that usually causes it.

In this one, specific case, those types of movements get put in the
tool path because you turned off the extra shells. You could leave the
extra shells on but then this part doesn't print correctly.

On different objects or maybe even on this object with a very
different profile config, you might get the problem only if you did
*not* turn extra shells off.

If you're not sure what extra shells are any why I suggested you turn
them off, I can go into that. It's good background/theory stuff to
know but it's not really required reading..

[ddurant]

> The bobble head would be funny but I think it would get old in like 5
> minutes.

You're probably right. Possibly after only 3 minutes.

They wouldn't *have to* be for the top of bots, though. You could put
them on pets or your car dash, too.

> > > - Show quoted text -- Hide quoted text -

[DiyAddicts]

>In this one, specific case, those types of movements get put in the
tool path because you turned off the extra shells.

Ok this makes a little more sense to me as to why it would behave like it is.

I'm at the point though to where I can not see anywhere that slipping could be happening. I'll try to shoot a video of the print maybe you can see something that I am not. I do have my travel rate up a little so I will bring that back down and try.

>If you're not sure what extra shells are any why I suggested you turn
them off, I can go into that. It's good background/theory stuff to
know but it's not really required reading..

I am actually trying to learn as much as I can about all this so more information is great to me! I looked for some more info on the extra shells and all I could really find was the two sentences on the Skeinforge wiki.

[ddurant]

A shell is basically the trace thread that the printer makes around
the object. When you print something, you usually see the printer draw
the outline then do lots of zig-zags to fill in the center then move
up and do the next layer. If you use extra shells, it draws more than
one outline around the object.

I think I've heard people say that extra shells adds strength but I'm
not sure I buy that. It seems like a nicely dialed-in profile should
have the same strength regardless of the shells. I could be wrong,
though.

Anyway, when you ask skeinforge to slice up an object, it looks at the
thread height and thread width that you specify (layer thickness and
layer thickness * w/t) then maps out the series of moves needed to
cover the object with threads of that size. Just like Crayons 101,
this is tracing the outline then scribbling to fill the area inside
the outline. Or, if you have extra shells, there are multiple *traces*
around the outline then scribbles.

One of the problems with doing things like this is that it's not
possible to fill areas that are between 2 and 3 times the thread
width. Not with skeinforge anyway.

If you haven't heard the 2-3 thing before, picture a Sharpie marker
that draws a line 1 inch wide. Using that marker and without drawing
over areas that you've already drawn on, can you trace & fill in an
square that's 2 inches wide? Yes - it's just two lines, side by side.
Just the traces around the outline cover the whole area.

A square 3 or more inches wide is also doable. You do the trace around
the edge then fill in the middle. The trace around the square takes up
2 inches then filling in the middle is just zig zagging through the
center.

An area that's more than 2 inches but less than 3 inches wide is a
different story, though. With that sort of area, you eat 2 inches
doing the trace around the outside but are then left with an area
that's less than 1 inch wide in the middle.

When you're using a marker, this really isn't that big of a deal - you
cheat and draw over an existing line. When you're extruding plastic,
it is a big deal. You can't just pump more plastic into the area
because it would blob up and cause all sorts of problems. You could
have software that looked for this sort of issue and did clever things
to decrease the thread width but, well, we don't have that software.

What skienforge does is leave that area blank. All your wobble
arrestor pictures (except the leaning one) show this issue - areas
that should be solid are left with big holes in the middle. Look
closely at just the top layer of the leaning one and compare it to the
tops of some of the others.

Once you sorta get that straight in your head (it took me a few tries
when I first heard about this), you can see why some objects are
sensitive to the number of extra shells: one extra trace around the
edge changes the 2-3 rule into a 4-5 rule because it eats up 2 more
threads widths (one on each side of the area being filled) and changes
where your limit is for filling in the center.

Not sure if that all makes sense - I tend to ramble - but I'm outta
time for now..

[Aaron Double]

Made sense to me:) Nice description.

[Aaron Double]

Sometimes, that space is important, especially if you want the part to
be springy. That's how the flexmount for the nut on the Z-rider works.

[Aaron Double]

Dave, if you go tonight, I'll have that part with me.

[ddurant]

> Made sense to me:)

You already know about this stuff and don't count!! :P

On Jan 6, 6:36 pm, Aaron Double <aad...@gmail.com> wrote:

> > > and all I could really find was the two sentences on the Skeinforge wiki.- Hide quoted text -

[DiyAddicts]

Wow! That was great. Couldn't have asked for a better explanation. I appreciate all your time helping with this. I'm getting a late start on things this evening, but I'm going look into things a little more. After a couple quick prints I can tell that the PID changes I made are making a noticeable difference.

One thing I have started noticing tonight is that when the HBP is heating is causes the DC extruder motor to drop pitch in sound, like it is slowing down some. Then as soon as the HPB led kicks off it kind of revs back up. It is almost impossible to tell how much it is actually slowing it down though.

I will have some more time this weekend to dive into things more. Again, thanks for all the help

[ddurant]

Glad it made sense..

Sorta the whole point was that the leaning you saw when you turned off
extra shells wasn't really about there being extra shells on the
object or not. Turning off the extra shells just left skeinforge
enough room to fill in the center of the arms on the object and the
moves needed to do that fill triggered the lean.

Your X or Y is indeed slipping. Maybe it only slips when you're doing
a difficult print like this one with lots of back & forth scribbles
but it is slipping and this will likely come back and bite you on some
other object. It could be that some other object slips when extra
shells are at their default settings. Or it could only happen when you
turn them up. Or down.

Feel free to ignore it for now if you want but keep in mind that
there's some threshold of some behavior that will make an axis slip. A
Cupcake is totally able to print that part correctly so I assume a ToM
can do it too. The trick will be figuring out what's causing the slip
then either how to fix it or how to avoid it in the future. This
doesn't need to be solved today but you should track it down and
squash it at some point.

> One thing I have started noticing tonight is that when the HBP is heating is
> causes the DC extruder motor to drop pitch in sound, like it is slowing down
> some

Wow. That's.. uh.. not very good. Hopefully MBI will have a solution
soon for all the people reporting motor/power troubles. A working
extruder isn't really an optional bit on a 3D printer.

I wonder if that wobble arrestor still slips with 0 shells if you
disable the HBP. Maybe the system sorta losing power making the
steppers more likely to slip steps...

[Ed Nisley]

On Fri, 2011-01-07 at 07:40 -0800, ddurant wrote:
> The trick will be figuring out what's causing the slip
> then either how to fix it or how to avoid it in the future.

Having fought this issue to a standstill on my Sherline CNC milling
machine, I think I know what's behind the "slipping" issue.

The problem is *not* that the belts slip on the pulleys. If that
happened, the extruder position would suddenly jump by an integer
multiple of the belt's 2 mm tooth pitch. An abrupt 2 mm offset in a
printed object would be painfully obvious.

What's really happening is that the extruder firmware doesn't
accelerate / decelerate the motors (as of the last time I looked at the
code, anyway). Instead, it assumes the motors can begin turning at
whatever step rate is required for the current motion and stop turning
when that pulse train ceases, with the carriage obediently following
along.

That's a good assumption for a machine with moderately beefy steppers,
good drivers, low-mass carriages, and low-friction/low-stiction guides.
Larger machines must ramp the motor / carriage velocity up and down,
limiting the acceleration at both ends of the motion to whatever the
hardware can handle.

The simplifying assumptions that permit bang-bang starts and stops are
true for the TOM (and Cupcake), except for friction / stiction. Despite
those lovely bronze bushings, it's entirely possible to build a TOM with
an X (or Y) carriage that the motors can move only with great
difficulty: I've done it!

The symptom is exactly what you're seeing: the commanded position
doesn't track the actual position, because the carriage doesn't start
moving until the motor manages to break it free from the stiction. Those
"lost steps" can never be made up, because the TOM doesn't know anything
about the actual carriage position.

In addition, the TOM stepper controllers use quarter-step drive, which
produces smoother motion at the cost of lower starting torque. Lower
torque = less force = lower maximum acceleration = less tolerance for
stiction.

Why it happens mostly in the X direction:

The X carriage (MBI calls it a "stage") slides on two rods, each of
which is fixed at four places: both ends of the Y carriage and both ends
of the X carriage. If the rod-to-rod spacing in those four places isn't
*exactly* equal, then the X carriage bushings will bind on the rods.
Alas, tolerance creep in the plywood and maybe a bit of off-center
sanding when you fitted the bushings into the plywood can produce
exactly that situation.

How to diagnose on X and Y:

Dismount the drive belt (which will probably require dismounting the
motors or their drive pulleys) so the carriage can slide without turning
the motor.

If you can bat the carriage back and forth along the entire length of
its rods by tapping (*not* pushing) with your fingertips, it's all good.

In my TOM, moving the X carriage required quite a shove and it
definitely didn't slide freely at all.

How to adjust the X carriage:

Take the entire X and Y assembly out of the TOM (sigh). Remove the rod
end caps on both ends of the rear rod. Slide the X carriage to the right
end, then push the rear rod out of the right end piece toward the left
(push it into the plywood) while supporting the X carriage. Most likely,
the rod will go *spung* a fraction of a millimeter horizontally (you
control the vertical offset by supporting the carriage).

That's the rod's way of telling you that the end hole is in the wrong
position. If it slides easily in and out of that hole, then it's all
good; go to the left end as below.

Sand the offending side of the hole (wrap sandpaper around a small rod
so it almost completely fills the hole) until the rod slides easily into
the hole.

Slide the carriage to the left end and repeat that process.

Replace both end caps.

The carriage should now slide back and forth with just finger taps.

If you overdo the sanding, shim the loose side of the holes with
aluminum foil and a dab of adhesive. If the rod rattles around, that's
bad; add an all-around shim and put a very thin slice of foam under the
end cap to calm it down.

How to adjust the Y carriage:

The right side of the Y carriage rides on bushings and the left side
rides on three ball bearings, making it not quite so sensitive to
misalignments. Check for free motion when you reinstall the assembly
and, if it's off, adjust the right-hand rod, because the ball bearings
don't capture the rod all the way around.

Those bronze bushings work wonderfully well, but only when the rods are
exactly parallel!

Hope that helps...

--
Ed
http://softsolder.com

[Aaron Double]

Hey DiyAddicts,

I printed some wobble arresters (not my arrester, natetrue's version,
same one you have been playing with) yesterday as calibration objects
for PLA.

I figured you might be interested. I am using a very modified cupcake
though, not a T-O-M.

That being said, it took me about 5months to start getting prints that
I was happy about (got my bot in Feburary 2010).

Just be patient, and you will be rewarded with good prints.

http://www.flickr.com/photos/48982649@N07/sets/72157625644168515/


I have a really dumb question, how do you upload an image into a post?
I haven't been able to find the instructions anywhere.

On Jan 6, 10:32 pm, DiyAddicts <adam_mar...@att.net> wrote:

[Koen Kooi]

Op 7 jan 2011, om 19:05 heeft Aaron Double het volgende geschreven:

> Hey DiyAddicts,
>
> I printed some wobble arresters (not my arrester, natetrue's version,
> same one you have been playing with) yesterday as calibration objects
> for PLA.
>
> I figured you might be interested. I am using a very modified cupcake
> though, not a T-O-M.
>
> That being said, it took me about 5months to start getting prints that
> I was happy about (got my bot in Feburary 2010).
>
> Just be patient, and you will be rewarded with good prints.
>
> http://www.flickr.com/photos/48982649@N07/sets/72157625644168515/
>
>
> I have a really dumb question, how do you upload an image into a post?
> I haven't been able to find the instructions anywhere.

This is a mailing list, so you can just attach it to the mail :)

regards,

koen

[Aaron Double]

Oh, I was getting too many emails from this so I turned that off. I
just look at it through the webpage. Maybe I need to adjust my
settings.

Thanks!

[ddurant]

Those came out great, Aaron!

I know you've got Robs stepper driver on your bot - did you go for
1/16th steps on X & Y? I did on mine but haven't installed it yet.
Ed's post about steppers has made me a little nervous..

> This is a mailing list, so you can just attach it to the mail :)

Unless you're not using it as a mailing list and using http://groups.google.com
instead! :P

On Jan 7, 1:11 pm, Koen Kooi <k...@beagleboard.org> wrote:

> koen- Hide quoted text -

[AARON DOUBLE]

Hey Dave,

Yes, I have Rob's PSMD running at 1/16th step, I feel like I have more torque with this over the makerbot old school steppers and I've been running it for about 5 hours today and the X stepper is not ever warm to the touch.

I also don't have the pots turned up. One of the interesting things that came out of this is the steppers are so quiet now that it's slightly musical, definitely different tones happening, I find it fun, my wife liked the louder buzzing more, easier for her to tune it out. I turned up the current a bit to make her happy.

[DiyAddicts]

> Sorta the whole point was that the leaning you saw when you turned off
extra shells wasn't really about there being extra shells on the
object or not. Turning off the extra shells just left skeinforge
enough room to fill in the center of the arms on the object and the
moves needed to do that fill triggered the lean.

Yeah sorry it took me so long to understand that :S

> Your X or Y is indeed slipping. Maybe it only slips when you're doing
a difficult print like this one with lots of back & forth scribbles
but it is slipping and this will likely come back and bite you on some
other object. It could be that some other object slips when extra
shells are at their default settings. Or it could only happen when you
turn them up. Or down.

I got the slipping problem fixed finally. I ended up swapping to a different power lead off the power supply and that seemed to do the trick. There was a very slight skip left after that, and bringing the travel rate down a hair got it the rest of the way.

photo1 attached attached is with 0 shells. There is not near enough plastic, but at least it printed without a lean!

>Wow. That's.. uh.. not very good. Hopefully MBI will have a solution
soon for all the people reporting motor/power troubles. A working
extruder isn't really an optional bit on a 3D printer.

Yeah I am definitely having some problems with power. After changing my PID settings the temp got a lot closer, went from about 19 degrees total temp swing to about 7-8 degrees.

I am still noticing my extruder slowing down (audibly) every time HBP kicks on to heat. So I figured I would turn it off and see what happened. Now with the HBP turned off the extruder still slows down but it much shorter burst. I moved some wires so I could see the extruder heater led, and now that I can see it, it shows that every time that heater kicks on it is also slowing down the extruder.

ANOTHER odd thing is when I have my HBP off my temp values on my extruder go up about 6 degrees and fluctuate from there. When I have it on my extruder temp barely gets above my set value (219)

So I am still trying with very little luck to get things dialed in. The one thing I don't know though is a 7-8 degree swing in extruder temperature too much? The second photo attached is of the dodecahedron, the one on the left is as close as I have been able to get. Right one is where I started out a couple weeks ago. The main flaw in the print is always in the lower half which you can see.

The settings on the left dodeca are:

Feed: 35
w/t:   1.5
LT:    .38
shells: 2
Temp: 219

I didn't get a picture of the top of it, but I can say it is a little sparse on top.

[DiyAddicts]

Wow that looks very nice, very jealous :)

> That being said, it took me about 5months to start getting prints that
I was happy about (got my bot in Feburary 2010).

I was hoping for more along the lines of 2 months. I knew this thing wouldn't print beautifully out of the box, so I was trying to be realistic (just maybe not realistic enough).

AM

[AARON DOUBLE]

About the bobble, it is coming from two things.

1. My build platform is now pretty heavy, so slinging it back and forth rapidly throws the whole makerbot around. It's a fairly large sized chunk of aluminum.

2. I'm now only doing one threaded rod. The center point on the nut on the threaded rod is at the same point (directly behind) as the tip of the mk5 nozzle. Since the mk5 is kinda tall and the weight is mostly on the top, when the makerbot is thrown around by really quick movements in the X-axis, the platform will rotationally vibrate with the nozzle tip as the center point, I actually designed it to do this. (I just didn't think it would work as well as it does)

The rotational vibration is the drawback of only having one vertical load point. I would love to know if anyone with a T-O-M sees this as well. I didn't notice it in action until recently because the prints weren't screwed up.

The other thing that is kinda built in is, since the platform is cantilevered from the back, there is a bit of vertical play that only shows up when there is an obstruction. In other words, if you have a really big blob the nozzle tip will just lift up a little bit and go over the blob and not tear the print off the platform. After a couple layers the blob is incorporated in the part. The blobs for me are because I have yet to put a Z endstop in place, if I screw up my starting height there is a chance of blobs.

On Jan 10, 2011, at 9:12 AM, coasterman-1674 wrote:

> If the X or Y is skipping, I would check belt tension, stepper
> parameters, and then adjust the Skeinforge settings until you can fit
> the extra shells in a dodecahedron so it draws with four lines. I
> think the goal here is to be able to smoosh 4 lines into 2mm of space.
> That will ease any zigzagging.
>
> And then the bobblehead makerbot, how funny. Make a Makerbot with a
> bobbling Z-stage. I laughed my head off seeing the next contest as a
> bobblehead as a joke. I'm not exactly sure if the Z should bobble like
> that, though. Mine never bobbles.

[ddurant]

I was actually just referring to the shaking my machine does when it
fills thin areas.

The part about the nut being level with the nozzle is interesting..
I'm wondering how that'll work with the makergear extruder. Also now
wondering about balance, since that stepper extruder has 90% of its
weight off to one side...

> > For more options, visit this group athttp://groups.google.com/group/makerbot?hl=en.- Hide quoted text -

[AARON DOUBLE]

The height is easily changable, either in the file or with some washers - shouldn't be a problem. About the balance, that also shouldn't be a problem, the vibration is always from where the platform is at in resting state. I'll do a weight test later today since I have one of those here as well.

[DiyAddicts]

> If the X or Y is skipping, I would check belt tension, stepper
parameters, and then adjust the Skeinforge settings until you can fit
the extra shells in a dodecahedron so it draws with four lines. I
think the goal here is to be able to smoosh 4 lines into 2mm of space.
That will ease any zigzagging.

It does not seem to be zigzagging when doing the fill.