Solidoodle WTF!
35 views
Skip to first unread message
Allen Brown
Sep 4, 2012, 6:16:14 PM9/4/12
to Reprap
A while back I ordered a Solidoodle 3D printer. When I learned
they wouldn't deliver until well after I would have paid the
credit card bill, making the charge uncontestable if they
*never* delivered, I canceled the order. They refunded
the charge to my card. That was a long time ago. Two months?
I'm not sure.
But today I was surprised when two boxes appeared on my doorstep.
A Solidoodle printer plus extra rolls of plastic fiber.
Now what? I'm inclined to power it up and try it out. If it
is good, I will pay them the regular price for it. If I don't
like it I will ship it back.
Note that I canceled that order based on their inappropriate
charge vs ship schedule. Not based on any impression of a
bad product. So now that the product is in my hands, I want
to see if it is good. And now that I don't have to risk paying
for something that is never delivered, I am potentially happy
to do business with them.
In the mean time, it is here. So if you want to see a
Solidoodle, come by Wednesday 6:30PM or so. I will finish
the unpacking and have it ready to test. Cassie, would you
like to help with the testing? Could be interesting.
--
Allen Brown http://brown.armoredpenguin.com/~abrown/
My $PID is Ingo Montoya. You kill -9 my Parent Process.
Prepare to VI! --- a bumper sticker
they wouldn't deliver until well after I would have paid the
credit card bill, making the charge uncontestable if they
*never* delivered, I canceled the order. They refunded
the charge to my card. That was a long time ago. Two months?
I'm not sure.
But today I was surprised when two boxes appeared on my doorstep.
A Solidoodle printer plus extra rolls of plastic fiber.
Now what? I'm inclined to power it up and try it out. If it
is good, I will pay them the regular price for it. If I don't
like it I will ship it back.
Note that I canceled that order based on their inappropriate
charge vs ship schedule. Not based on any impression of a
bad product. So now that the product is in my hands, I want
to see if it is good. And now that I don't have to risk paying
for something that is never delivered, I am potentially happy
to do business with them.
In the mean time, it is here. So if you want to see a
Solidoodle, come by Wednesday 6:30PM or so. I will finish
the unpacking and have it ready to test. Cassie, would you
like to help with the testing? Could be interesting.
--
Allen Brown http://brown.armoredpenguin.com/~abrown/
My $PID is Ingo Montoya. You kill -9 my Parent Process.
Prepare to VI! --- a bumper sticker
Shannon Dealy
Sep 5, 2012, 2:52:28 AM9/5/12
to Reprap
[snip]
> In the mean time, it is here. So if you want to see a
> Solidoodle, come by Wednesday 6:30PM or so. I will finish
> the unpacking and have it ready to test. Cassie, would you
[snip]
> Solidoodle, come by Wednesday 6:30PM or so. I will finish
> the unpacking and have it ready to test. Cassie, would you
I can't make it, but will be interested to hear how it goes.
Shannon C. Dealy | DeaTech Research Inc.
de...@deatech.com | - Custom Software Development -
Phone: (800) 467-5820 | - Natural Building Instruction -
or: (541) 929-4089 | www.deatech.com
abr...@peak.org
Sep 6, 2012, 12:45:17 AM9/6/12
to corvallis...@googlegroups.com
Jeremy joined me for this session.
Solidoodle says to use pronterface software. I did.
Temperature for head and bed worked. Oddly enough
their (online) instructions specify temperatures to
use but don't say what kind of plastic their
instructions are for. Not knowing any better back
when I ordered the printer, I asked for ABS and that
is what was included. So that's what I tested with.
The Y axis moved and homed well. The X axis moved
but would not home. The Z axis would only home if
I first moved Z to very close to the stop. Otherwise
it would move a short distance and then scream for a
long time like the motor was jammed. Sometimes when
I told the printer to move Y +1mm, it would instead
move -X some large distance and jam it against the far
stop. These may be pronterface bugs for all I know.
I was never particularly successful driving the old
reprap from my laptop.
With head at 200C and bed about 90C, I printed a
cube. It had trouble adhering the first time.
After I adjusted the Z stop, it did much better.
But the bed doesn't seem to be flat. One corner
didn't adhere. And the plastic is maximally flat
at the opposite corner where it did adhere.
The walls of the cube are not dimensionally
consistent. But the cube appeared to be moving
on the bed where it was poorly attached. Possibly
if I properly level the bed it will work better.
Notably, Solidoodle says you may have to level
the bed but doesn't give instructions for how
to do this.
--
Allen Brown abrown at peak.org http://brown.armoredpenguin.com/~abrown/
"... one of the main causes of the fall of the Roman Empire
was that, lacking zero, they had no way to indicate successful
termination of their C programs." --- Robert Firth
--
You received this message because you are subscribed to the Google Groups "corvallis-techshop" group.
To post to this group, send email to corvallis...@googlegroups.com.
To unsubscribe from this group, send email to corvallis-techs...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/corvallis-techshop?hl=en.
Solidoodle says to use pronterface software. I did.
Temperature for head and bed worked. Oddly enough
their (online) instructions specify temperatures to
use but don't say what kind of plastic their
instructions are for. Not knowing any better back
when I ordered the printer, I asked for ABS and that
is what was included. So that's what I tested with.
The Y axis moved and homed well. The X axis moved
but would not home. The Z axis would only home if
I first moved Z to very close to the stop. Otherwise
it would move a short distance and then scream for a
long time like the motor was jammed. Sometimes when
I told the printer to move Y +1mm, it would instead
move -X some large distance and jam it against the far
stop. These may be pronterface bugs for all I know.
I was never particularly successful driving the old
reprap from my laptop.
With head at 200C and bed about 90C, I printed a
cube. It had trouble adhering the first time.
After I adjusted the Z stop, it did much better.
But the bed doesn't seem to be flat. One corner
didn't adhere. And the plastic is maximally flat
at the opposite corner where it did adhere.
The walls of the cube are not dimensionally
consistent. But the cube appeared to be moving
on the bed where it was poorly attached. Possibly
if I properly level the bed it will work better.
Notably, Solidoodle says you may have to level
the bed but doesn't give instructions for how
to do this.
--
Allen Brown abrown at peak.org http://brown.armoredpenguin.com/~abrown/
"... one of the main causes of the fall of the Roman Empire
was that, lacking zero, they had no way to indicate successful
termination of their C programs." --- Robert Firth
You received this message because you are subscribed to the Google Groups "corvallis-techshop" group.
To post to this group, send email to corvallis...@googlegroups.com.
To unsubscribe from this group, send email to corvallis-techs...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/corvallis-techshop?hl=en.
Kassandra Kaplan
Sep 6, 2012, 1:02:47 AM9/6/12
to corvallis...@googlegroups.com, corvallis...@googlegroups.com
For abs, the temps seem low. Normally, abs is 200-220 and the bed is 90-120. Assuming even heating, bed leveling would explain the sticking problem. If the bed is unevenly heated increasing the temp by 5c can help.
Not sure about the z axis, my guess is that it might need some lubrication.
Anyway, I can not do Wednesdays for a while.
どうぞおげんきで
Kassie
Not sure about the z axis, my guess is that it might need some lubrication.
Anyway, I can not do Wednesdays for a while.
どうぞおげんきで
Kassie
abr...@peak.org
Sep 6, 2012, 11:00:21 AM9/6/12
to corvallis...@googlegroups.com
----- "Kassandra Kaplan" <shado...@gmail.com> wrote:
> For abs, the temps seem low. Normally, abs is 200-220 and
> the bed is 90-120. Assuming even heating, bed leveling
> would explain the sticking problem. If the bed is
> unevenly heated increasing the temp by 5c can help.
Solidoodle warns against raising the head temp above 200.
> For abs, the temps seem low. Normally, abs is 200-220 and
> the bed is 90-120. Assuming even heating, bed leveling
> would explain the sticking problem. If the bed is
> unevenly heated increasing the temp by 5c can help.
Evidently their heater can't handle it.
The bed struggled to reach 100.
> Not sure about the z axis, my guess is that it might need some
> lubrication.
the button to move the Z axis manually, it never
complained. I think it is a software problem.
> Anyway, I can not do Wednesdays for a while.
>
> どうぞおげんきで
> Kassie
A modest man, who has much to be modest about (On Clement Atlee)
--- Sir Winston Churchill
Carl Price
Sep 6, 2012, 11:52:49 AM9/6/12
to corvallis...@googlegroups.com
Allen, do you have any PLA you can try? If not let me know and I can bring you some to try out.
Working through the issues on my printer, bed level is critical, I would like to see the ability to level, even if it is just measure and report, built in, it seems to be the biggest variable on printing well.
Anyway, let me know on the PLA, I can grab some off a spool and drop it after work some night (will need a day so I can bring it with me to work).
Kassandra Kaplan
Sep 6, 2012, 12:57:58 PM9/6/12
to corvallis...@googlegroups.com, corvallis...@googlegroups.com
Pla uses more forgiving temps, 185-200 for the hotend and 50-60 for the bed, but is more picky to get perfect results.
Only 200!! I bet it can go higher but it's pushing the material used. The big concern is ptfe that off gas and softens/melts around 240, but most adhesives/sealants can not handle it. Lastly, the may be trying to call it safe since most FFD machines these day use over powered resistors.
I have gotten good results at 80c and 200 for abs, but sticking was not guaranteed. So those setting should be workable, but bed leveling is critical!
I'll bring some 1.75 and 3mm pla when I come out to help out.
I'll bring some 1.75 and 3mm pla when I come out to help out.
どうぞおげんきで
Kassie
abr...@peak.org
Sep 6, 2012, 2:04:55 PM9/6/12
to corvallis...@googlegroups.com
I would have used PLA if I had some. I understand it is
easier to use.
I plan to level it today. Perhaps in an hour. I haven't
tried yet, but it doesn't look difficult.
Switching from ABS to PLA without changing over the printhead
may or may not work. Is PLA safe at 200C? I think it would
char at 220C. And the ABS likely wouldn't flush out of the
head at lower than 200C. So if it is OK at 200C, then this
seems OK.
Tomorrow would be good. And you can look over the Solidoodle.
BTW, how is your printer doing?
If a mouse lives in the cookie jar, that does not necessarily make
him a cookie. --- from The Hiding Place
easier to use.
I plan to level it today. Perhaps in an hour. I haven't
tried yet, but it doesn't look difficult.
Switching from ABS to PLA without changing over the printhead
may or may not work. Is PLA safe at 200C? I think it would
char at 220C. And the ABS likely wouldn't flush out of the
head at lower than 200C. So if it is OK at 200C, then this
seems OK.
Tomorrow would be good. And you can look over the Solidoodle.
BTW, how is your printer doing?
If a mouse lives in the cookie jar, that does not necessarily make
him a cookie. --- from The Hiding Place
abr...@peak.org
Sep 6, 2012, 2:21:20 PM9/6/12
to corvallis...@googlegroups.com
OK, so I should be able to safely flush ABS using PLA at
200C. Good to know.
This printer uses 1.75mm filament.
A man who surrenders when he's wrong, is honest. A man who
surrenders when he's not sure is wise. A man who surrenders
even if he's right, is a husband. --- Author Unknown
----- Original Message -----
From: "Kassandra Kaplan" <shado...@gmail.com>
To: corvallis...@googlegroups.com
Cc: corvallis...@googlegroups.com
Sent: Thursday, September 6, 2012 9:57:58 AM
Subject: Re: Solidoodle WTF!
200C. Good to know.
This printer uses 1.75mm filament.
A man who surrenders when he's wrong, is honest. A man who
surrenders when he's not sure is wise. A man who surrenders
even if he's right, is a husband. --- Author Unknown
----- Original Message -----
From: "Kassandra Kaplan" <shado...@gmail.com>
To: corvallis...@googlegroups.com
Cc: corvallis...@googlegroups.com
Sent: Thursday, September 6, 2012 9:57:58 AM
Subject: Re: Solidoodle WTF!
Kassandra Kaplan
Sep 6, 2012, 2:27:35 PM9/6/12
to corvallis...@googlegroups.com, corvallis...@googlegroups.com
Short answer, No problem flushing the head.
Long answer, Abs is capable of extruding under 190, but it will not stick well and is too stiff to extrude fast. Pla can get dicey above 215 by causing jam if it sits, but I have been able to run pla constantly at 225. At 200 you could leave your hotend on overnight not extruding and only get the slightest blockage that clear quickly with normal usage. This means their is a lot of overlap in tempurate. In fact you may be able to use normal pla temps to clear.
Additionally the FFD hotends have very laminar flows, so the should clear quickly and thoughly with about 100mm of the new material. I normally like to extrude 20mm wait for some mixing and do 5-10mm then wait cycles at a time till clear.
Btw, Monday evening does work for me.
どうぞおげんきで
Kassie
Long answer, Abs is capable of extruding under 190, but it will not stick well and is too stiff to extrude fast. Pla can get dicey above 215 by causing jam if it sits, but I have been able to run pla constantly at 225. At 200 you could leave your hotend on overnight not extruding and only get the slightest blockage that clear quickly with normal usage. This means their is a lot of overlap in tempurate. In fact you may be able to use normal pla temps to clear.
Additionally the FFD hotends have very laminar flows, so the should clear quickly and thoughly with about 100mm of the new material. I normally like to extrude 20mm wait for some mixing and do 5-10mm then wait cycles at a time till clear.
Btw, Monday evening does work for me.
どうぞおげんきで
Kassie
Carl Price
Sep 6, 2012, 2:47:10 PM9/6/12
to corvallis...@googlegroups.com
I have two spools of pla, but only 3mm, sounds like it may not work (let me know if it will).
My printer is finally starting to get decent prints. I did a box for a board stack that worked well, I now need to grab a couple models and run some through to see if it will reliably do good prints.
--Carl
Carl Price
Sep 6, 2012, 2:54:03 PM9/6/12
to corvallis...@googlegroups.com
Looked at the website for it, looks like a nice little printer. One thing I have liked is the blue painters tape over kapton. It isn't supposed to last as long, but I find I get good adhesion to it, and it's cheaper, so doesn't bother me to replace a little more often. Home depot has the wide rolls too.
abr...@peak.org
Sep 6, 2012, 3:01:48 PM9/6/12
to corvallis...@googlegroups.com
I doubt 3mm will fit.
Don't ever take a fence down until you know why it was put up.
--- Robert Frost
Don't ever take a fence down until you know why it was put up.
--- Robert Frost
abr...@peak.org
Sep 6, 2012, 7:45:53 PM9/6/12
to corvallis...@googlegroups.com
The bed took 20 minutes to reach 100C.
Today the Z axis behaved correctly.
The Y axis continues to fail to home.
The bed is not flat. The center is higher
than the edges. I printed another cube.
Good adhesion to the bed. And it appeared
flat on the bed. The rest of the cube is
still pretty sloppy.
Few things are harder to put up with than the
annoyance of a good example. ---Mark Twain
----- Original Message -----
From: abr...@peak.org
To: corvallis...@googlegroups.com
Sent: Thursday, September 6, 2012 8:00:21 AM
Subject: Re: Solidoodle WTF!
Today the Z axis behaved correctly.
The Y axis continues to fail to home.
The bed is not flat. The center is higher
than the edges. I printed another cube.
Good adhesion to the bed. And it appeared
flat on the bed. The rest of the cube is
still pretty sloppy.
Few things are harder to put up with than the
annoyance of a good example. ---Mark Twain
----- Original Message -----
From: abr...@peak.org
To: corvallis...@googlegroups.com
Sent: Thursday, September 6, 2012 8:00:21 AM
Subject: Re: Solidoodle WTF!
abr...@peak.org
Sep 9, 2012, 3:51:40 PM9/9/12
to corvallis...@googlegroups.com
Kassie,
Are we still on for Monday evening?
Anybody else want to see the Solidoodle Monday?
An expert is just a former drip under pressure.
----- Original Message -----
From: "Kassandra Kaplan" <shado...@gmail.com>
To: corvallis...@googlegroups.com
Are we still on for Monday evening?
Anybody else want to see the Solidoodle Monday?
An expert is just a former drip under pressure.
----- Original Message -----
From: "Kassandra Kaplan" <shado...@gmail.com>
To: corvallis...@googlegroups.com
Kassandra Kaplan
Sep 10, 2012, 5:49:29 PM9/10/12
to corvallis...@googlegroups.com, corvallis...@googlegroups.com
Planning for 6:30pm.
どうぞおげんきで
Kassie
どうぞおげんきで
Kassie
abr...@peak.org
Sep 10, 2012, 6:06:52 PM9/10/12
to corvallis...@googlegroups.com
Great! Bring your laptop.
And it might be interesting to bring the spare electronics.
Cauliflower is nothing but cabbage with a college education.
--- Mark Twain
And it might be interesting to bring the spare electronics.
Cauliflower is nothing but cabbage with a college education.
--- Mark Twain
abr...@peak.org
Sep 11, 2012, 2:17:39 AM9/11/12
to corvallis...@googlegroups.com
Kassie got the same results on the Z and Y
axis home attempts. It's starting to look
like something internal to the Solidoodle.
I just realized the cubes I had printed last
time were done with parameters for the old
reprap with 3mm filament. We will have to
try it with a regenerated file.
Kassie noticed that the Z axis home motion rate
was too fast. She halved the number and it
started working correctly.
She went online and found someone reporting
exactly the problems we discovered. That guy
also halved the Z axis rate. And he determined
that the Y axis home was in the wrong direction.
She switched that in the Solidoodle and Y
axis began homing.
Solidoodle really should have righted these
problems before shipping the first unit. Not
impressive.
Kassie adjusted some parameters. It's working
better. There are still some problem areas.
Realization: they are not dealing with backlash.
Especially with the Z axis, this may be biting
us. Considering writing a filter to adjust
all Z movements in the gcode.
After 6 passes of adjusting things it is pretty
good. I think I'll keep it.
Things to do:
1. filament feed needs to be fixed so it doesn't tangle
2. I need to take Kassies settings and move them
to Linux and get it working here.
3. Write gcode filter to fix backlash problem with Z.
4. File bug report with slic3r about backlash,
including examples from 3.
There's a perfectly good explanation for this, which
I'll make up later. --- Bug Hall in "The Little Rascals"
----- Original Message -----
From: abr...@peak.org
To: corvallis...@googlegroups.com
Sent: Thursday, September 6, 2012 4:45:53 PM
Subject: Re: Solidoodle WTF!
axis home attempts. It's starting to look
like something internal to the Solidoodle.
I just realized the cubes I had printed last
time were done with parameters for the old
reprap with 3mm filament. We will have to
try it with a regenerated file.
Kassie noticed that the Z axis home motion rate
was too fast. She halved the number and it
started working correctly.
She went online and found someone reporting
exactly the problems we discovered. That guy
also halved the Z axis rate. And he determined
that the Y axis home was in the wrong direction.
She switched that in the Solidoodle and Y
axis began homing.
Solidoodle really should have righted these
problems before shipping the first unit. Not
impressive.
Kassie adjusted some parameters. It's working
better. There are still some problem areas.
Realization: they are not dealing with backlash.
Especially with the Z axis, this may be biting
us. Considering writing a filter to adjust
all Z movements in the gcode.
After 6 passes of adjusting things it is pretty
good. I think I'll keep it.
Things to do:
1. filament feed needs to be fixed so it doesn't tangle
2. I need to take Kassies settings and move them
to Linux and get it working here.
3. Write gcode filter to fix backlash problem with Z.
4. File bug report with slic3r about backlash,
including examples from 3.
There's a perfectly good explanation for this, which
I'll make up later. --- Bug Hall in "The Little Rascals"
----- Original Message -----
From: abr...@peak.org
To: corvallis...@googlegroups.com
Sent: Thursday, September 6, 2012 4:45:53 PM
Subject: Re: Solidoodle WTF!
The bed took 20 minutes to reach 100C.
Today the Z axis behaved correctly.
The Y axis continues to fail to home.
The bed is not flat. The center is higher
than the edges. I printed another cube.
Good adhesion to the bed. And it appeared
flat on the bed. The rest of the cube is
still pretty sloppy.
Today the Z axis behaved correctly.
The Y axis continues to fail to home.
The bed is not flat. The center is higher
than the edges. I printed another cube.
Good adhesion to the bed. And it appeared
flat on the bed. The rest of the cube is
still pretty sloppy.
Few things are harder to put up with than the
annoyance of a good example. ---Mark Twain
Bob Miller
Sep 11, 2012, 12:49:21 PM9/11/12
to corvallis...@googlegroups.com
Can you explain the backlash a little more?
What exactly was it doing wrong?
Thanks.
Bob Miller K<bob>
kb...@jogger-egg.com
What exactly was it doing wrong?
Thanks.
kb...@jogger-egg.com
abr...@peak.org
Sep 11, 2012, 2:50:56 PM9/11/12
to corvallis...@googlegroups.com
Backlash happens in pretty much all mechanical
systems when there is a mechanical linkage
between the measurement and drive point and
the thing you are controlling. The classic case
is a milling machine. You turn a graduated dial
which drives gears which move the table under
the cutting tool. Specifically backlash happens
when you change directions. If you are cranking
the table to the left and then switch to the right
you have to crank it perhaps an 1/8 turn before
the table responds.
The reprap and solidoodle design have this situation
with the Z axis. Actually all the axes have it, but
Z is particularly bad because it is screw driven.
The table is driven up and down by a screw (and possibly
belts and gears). The stepper motor both drives
the motion and measures it. But it is the table
position we care about. And that is decoupled by
all that mechanical crap in between.
When you home the printer, it raises the table to
the print head. The last motion is always moving
the table up. You can't go up any more because
you would jam the print head. And you can't go
down (yet) because you need that tight clearance
for the first layer. Once you finish the first layer,
the printer lowers the table. But now we are moving
it down. Backlash! It doesn't move down as far as
the stepper says it does.
Machinists have known about this for centuries. They
came up with a very effective solution long long ago.
You pick a direction. All measurements are done after
moving in that direction.
Since the printer has no choice but to be moving the
table up after a home, that is the only choice we have
for a measurement direction. We should always move
the table up. Well obviously we have to be able to move
it down. So we should do what machinists do. Go
further than we need in the wrong direction and then
approach the right location from the right direction,
in this case upward.
Let's say we want to drop the table 0.3mm. To solve
the backlash problem, we drop the table 1.3mm and then
raise it 1mm. That eliminates any backlash that is
less than 1mm. Online reports say the backlash in
the solidoodle is less than 0.2mm. So 1mm should be
more than enough.
This should be easy to filter into the gcode, since
that is a text file. I'm not much of a programmer,
but I'm sure I can bang this out in awk.
You can tell you have backlash in your printer if the
first two layers are different from all the others.
Peace comes from within. Do not seek it without. --- Buddha
systems when there is a mechanical linkage
between the measurement and drive point and
the thing you are controlling. The classic case
is a milling machine. You turn a graduated dial
which drives gears which move the table under
the cutting tool. Specifically backlash happens
when you change directions. If you are cranking
the table to the left and then switch to the right
you have to crank it perhaps an 1/8 turn before
the table responds.
The reprap and solidoodle design have this situation
with the Z axis. Actually all the axes have it, but
Z is particularly bad because it is screw driven.
The table is driven up and down by a screw (and possibly
belts and gears). The stepper motor both drives
the motion and measures it. But it is the table
position we care about. And that is decoupled by
all that mechanical crap in between.
When you home the printer, it raises the table to
the print head. The last motion is always moving
the table up. You can't go up any more because
you would jam the print head. And you can't go
down (yet) because you need that tight clearance
for the first layer. Once you finish the first layer,
the printer lowers the table. But now we are moving
it down. Backlash! It doesn't move down as far as
the stepper says it does.
Machinists have known about this for centuries. They
came up with a very effective solution long long ago.
You pick a direction. All measurements are done after
moving in that direction.
Since the printer has no choice but to be moving the
table up after a home, that is the only choice we have
for a measurement direction. We should always move
the table up. Well obviously we have to be able to move
it down. So we should do what machinists do. Go
further than we need in the wrong direction and then
approach the right location from the right direction,
in this case upward.
Let's say we want to drop the table 0.3mm. To solve
the backlash problem, we drop the table 1.3mm and then
raise it 1mm. That eliminates any backlash that is
less than 1mm. Online reports say the backlash in
the solidoodle is less than 0.2mm. So 1mm should be
more than enough.
This should be easy to filter into the gcode, since
that is a text file. I'm not much of a programmer,
but I'm sure I can bang this out in awk.
You can tell you have backlash in your printer if the
first two layers are different from all the others.
Peace comes from within. Do not seek it without. --- Buddha
Brian H Wilson
Sep 11, 2012, 2:59:37 PM9/11/12
to corvallis...@googlegroups.com
> This should be easy to filter into the gcode, since
> that is a text file. I'm not much of a programmer,
> but I'm sure I can bang this out in awk.
Wow - I am old enough to know what awk is,
> that is a text file. I'm not much of a programmer,
> but I'm sure I can bang this out in awk.
but too young to know how to use it.
I jumped in when perl was the thing.
I do have the sed & awk t-shirt though.
I'd bang it out in python these days. Then again since I still work as a
programmer they make me learn new languages. It's Java and Android this
week.
BTW someone gave me an MG a few weeks ago so I am currently inclined to
build out my own shop right now vs trying to get a maker space going.
Too bad, there must be 50 private shops already spread around Corvallis.
Maybe we should work out a map and a list of who has what available
instead of trying to get it all under one roof.
Brian
Kassandra Kaplan
Sep 11, 2012, 3:14:09 PM9/11/12
to corvallis...@googlegroups.com
The z axis is homed by moving the platform up to the extruder. It then prints the first layer and drops the platform. This drop consistently ends up being too large and the next layer is too far from the first layer to adhere.
I hacked together a fudge factor in slic3r but it is not ideal because it makes the gcode more machine specific and is creating a theoretical .2mm z dimensional error ( in addition to the .5mm normal dimensional error i would expect).
I'm not fully convinced this is really backlash. I would expect backlash to put these two layers too close together, do to a lack of responsiveness in direction changes. However this seem to be like backlash.
In standard reprap, backlash is handled as a mechanical problem. Normally the backlash is such a small factor (compared to all the other errors) that it is ignored and because these machines are viewed as hand build, backlash is expect to be widely varied such that a default expected backlash would not be useful. In short, for most repraps backlash software solutions are not a priority compared to dual-extruder support, fans, speed, temp stabilization, better slicing algorithms, same material support, look ahead for smoothing, etc. The problem of backlash is understood, but not dealt with yet.
どうぞおげんきで
Kassie
I hacked together a fudge factor in slic3r but it is not ideal because it makes the gcode more machine specific and is creating a theoretical .2mm z dimensional error ( in addition to the .5mm normal dimensional error i would expect).
I'm not fully convinced this is really backlash. I would expect backlash to put these two layers too close together, do to a lack of responsiveness in direction changes. However this seem to be like backlash.
In standard reprap, backlash is handled as a mechanical problem. Normally the backlash is such a small factor (compared to all the other errors) that it is ignored and because these machines are viewed as hand build, backlash is expect to be widely varied such that a default expected backlash would not be useful. In short, for most repraps backlash software solutions are not a priority compared to dual-extruder support, fans, speed, temp stabilization, better slicing algorithms, same material support, look ahead for smoothing, etc. The problem of backlash is understood, but not dealt with yet.
どうぞおげんきで
Kassie
Brian H Wilson
Sep 11, 2012, 3:15:01 PM9/11/12
to corvallis...@googlegroups.com
Getting tired of one subject line. :-)
On 9/11/2012 11:50 AM, abr...@peak.org wrote:
> Backlash happens in pretty much all mechanical
> systems when there is a mechanical linkage
This is one of the reasons you can't buy a cheapo milling machine for a
few $K and then throw encoders and stepper motors on it.
CNC machines use things like ball bearings in the lead screws to take
out the play. See http://en.wikipedia.org/wiki/Ball_screw
You can take out the play with software, sure, but that means you can
never change direction in the middle of a cut (or in a printer I guess
that would be the middle of a "paste".) so you can't make things with
complex curves.
BTW when I got that MG it was in a barn we were tearing down that had a
lead screw driven door opener so I grabbed that too. I had to cut the
screw to get it home in the Yaris so now I have TWO lead screws :-) each
about 2 meters long and one heavy duty motor drive with built in RF
remote control!
Everybody else saw yet another piece of junk to get rid of, I saw a big
X-Y drive system (or maybe it's a Z drive). This explains the confused
crowded state of my garage and my "lab". Julie wonders sometimes what
would happen to it all if I were to pass away.
Brian
On 9/11/2012 11:50 AM, abr...@peak.org wrote:
> Backlash happens in pretty much all mechanical
> systems when there is a mechanical linkage
few $K and then throw encoders and stepper motors on it.
CNC machines use things like ball bearings in the lead screws to take
out the play. See http://en.wikipedia.org/wiki/Ball_screw
You can take out the play with software, sure, but that means you can
never change direction in the middle of a cut (or in a printer I guess
that would be the middle of a "paste".) so you can't make things with
complex curves.
BTW when I got that MG it was in a barn we were tearing down that had a
lead screw driven door opener so I grabbed that too. I had to cut the
screw to get it home in the Yaris so now I have TWO lead screws :-) each
about 2 meters long and one heavy duty motor drive with built in RF
remote control!
Everybody else saw yet another piece of junk to get rid of, I saw a big
X-Y drive system (or maybe it's a Z drive). This explains the confused
crowded state of my garage and my "lab". Julie wonders sometimes what
would happen to it all if I were to pass away.
Brian
Bob Miller
Sep 11, 2012, 3:36:15 PM9/11/12
to corvallis...@googlegroups.com
Thanks, all, for the explanations.
Reply all
Reply to author
Forward
0 new messages