Crooked Printer
64 views
Skip to first unread message
Ryan Carlyle
Apr 7, 2019, 1:07:17 AM4/7/19
to 3DP Ideas
Thinking about a show-printer that is easier to travel with than my Tesseract Delta. I could pick back up my collapsing printer (gantry extends up in Z while it prints) or I have a seed or an idea for a printer that is built crooked and janky (or even different every time you assemble it) but still prints straight.
There are mechanical options for running on irregular shapes. For example, if you build any four-sided irregular polygon, the midpoints of each side always form a parallelogram. So you can go from a wobbly-looking trapezoid to having nice parallel sides for a rhombus drive firmware mode printer or whatever. Just need a self-centering mechanism to support linear hardware and a drive mechanism that is somewhat length-agnostic. (Screws can overshoot the working volume to do this.) Complex but very designable.
Another option is to use a skew correction scheme from the Prusa Marlin fork. Or grab dimensions via min+max homing with carriage mounted endstop switches and calculate build area shape from that. I’m not opposed to doing some light firmware work on this.
I’m still soaking on options here.
There are mechanical options for running on irregular shapes. For example, if you build any four-sided irregular polygon, the midpoints of each side always form a parallelogram. So you can go from a wobbly-looking trapezoid to having nice parallel sides for a rhombus drive firmware mode printer or whatever. Just need a self-centering mechanism to support linear hardware and a drive mechanism that is somewhat length-agnostic. (Screws can overshoot the working volume to do this.) Complex but very designable.
Another option is to use a skew correction scheme from the Prusa Marlin fork. Or grab dimensions via min+max homing with carriage mounted endstop switches and calculate build area shape from that. I’m not opposed to doing some light firmware work on this.
I’m still soaking on options here.
Whosawhatsis
Apr 7, 2019, 1:50:27 AM4/7/19
to Ryan Carlyle, 3DP Ideas
This thread may also be of interest: https://forum.makerforums.info/t/infinite-z-printer-brainstorming/74930/38
--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To post to this group, send email to 3dp-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/b86c6e8c-c5f9-469d-ab03-db50f610bcf4%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Ryan Carlyle
Apr 7, 2019, 8:03:37 PM4/7/19
to 3DP Ideas
Good read. Nick and I build very different things though. I like linearizing in hardware, he likes using firmware... night and day!
Whosawhatsis
Apr 7, 2019, 9:57:34 PM4/7/19
to Ryan Carlyle, 3DP Ideas
I didn't imagine you would want to use the same solution that Nick would come up with, I just wanted to point out a parallel discussion with similar(ish) goals.
Btw, I don't know if you've logged into that site yet, but they've ported the posts from a bunch of the G+ communities over there, and if you use your google account to log in, your old posts will be linked with your account.
On April 7, 2019 at 17:03:38, Ryan Carlyle (temp...@gmail.com) wrote:
Good read. Nick and I build very different things though. I like linearizing in hardware, he likes using firmware... night and day!
--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To post to this group, send email to 3dp-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/43cf1399-dad3-4362-862d-eb18617f3102%40googlegroups.com.
dan...@puptv.com
Apr 7, 2019, 10:45:43 PM4/7/19
to 3DP Ideas
I think one contender for this may be the unfinished Flying SkyDelta with Super Gravity Pole...
It starts with a triangular anti-prism: (Two horizontal triangles at the top and bottom point different directions, with each apex connected by poles going up at an angle.)
This structure, even with rather light poles is incredibly strong, and stable.
The Flying SkyDelta is a winch based delta printer.
and The Super-Gravity Pole. If you just had an effector supported by winches, it would need to be huge (to always have a very stable resting point), heavy, and move very slowly. By creating a downward force on the effector, you can start printing incredibly quickly...
Here's a short video clip: https://www.youtube.com/watch?v=AgsFymFxr7g
The advantage of this design is that it can collapse to an electronics box and a single tube with 13 sticks (3 bottom, 6 angled, 3 top, and 1 super gravity pole).
The disadvantage with the design is that string based printers are troublesome to string together... This could be made easier with an electronics program where the motion motors keep a constant tension on the lines. (I wonder if running a program to constantly tension stepper motors with a very low current would work to make it easier.)
Alternately you may want to consider the hanging printer:
Daniel - www.TriDPrinting.com/
Ryan Carlyle
Apr 8, 2019, 9:30:42 AM4/8/19
to 3DP Ideas
The SkyDelta would be a really good show printer, yeah. You got it printing, didn't you?
Ryan Carlyle
Apr 8, 2019, 9:31:42 AM4/8/19
to 3DP Ideas
I have been meaning to get onto Makerforums, it's just not in my regular rotation. Most of my 3DP social media browsing is via my phone while waiting on my kids to fall asleep so if there's no app I'm a lot less likely to look at it. Lemme go post now...
dan...@puptv.com
Apr 8, 2019, 10:52:52 AM4/8/19
to 3DP Ideas
The Skydelta did print, but the calibration/kinematics never quite worked correctly.
Too many assumptions, some of which were not correct enough to work were used in the software.
For example, I used small v bearings instead of a small eyelet at the top corners, yet assumed that I could find an optimal virtual "location" for an eyelet that would work well enough.
In the end, it seemed that a cmm would be necessary to get accurate location feedback and determine more accurate kinematic functions.
Daniel
Whosawhatsis
Apr 8, 2019, 3:08:02 PM4/8/19
to Ryan Carlyle, 3DP Ideas
Add to home screen works, at least on android. I can even get push notifications of replies.
On April 8, 2019 at 06:31:43, Ryan Carlyle (temp...@gmail.com) wrote:
I have been meaning to get onto Makerforums, it's just not in my regular rotation. Most of my 3DP social media browsing is via my phone while waiting on my kids to fall asleep so if there's no app I'm a lot less likely to look at it. Lemme go post now...
--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To post to this group, send email to 3dp-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/40492444-0019-46da-aac8-fe49ef8b82de%40googlegroups.com.
Ryan Carlyle
Apr 8, 2019, 3:15:56 PM4/8/19
to 3DP Ideas
Anybody know of any existing firmware capabilities for the printer to self-calibrate how big its printing envelope is? Seems like an easy thing to accomplish with stall detect drivers and a Z probe. I'd like to be able to make a printer with a variable size build volume and have it find its own center and home positions.
Ryan Carlyle
Apr 8, 2019, 9:36:11 PM4/8/19
to 3DP Ideas
What I’m thinking right now is an XY gantry build on a non-rectangular frame. It will have four different frame side lengths and at least one will be adjustable length / telescoping. So i’ll take the printer out of a backpack, unfold it to some random crooked orientation, lock it in place, and run a self-calibrate.
Each of the four XY crooked frame members will have an attachment point at the center of the beam. On the adjustable length beam(s), a self-centering double gear rack will provide the center point. Four beam centers produces a parallelogram. Then linear rods are run between attachment points to produce a parallel mount for a gantry. The gantry bridge rods extend past one carriage to accommodate the differing parallelogram width. Both X and Y run on screws which can also extend past the gantry. So now we have a working XY gantry mounted on an arbitrary-shaped frame.
Then mount non-contact switches at MID span for XY homing, set switches as 0,0, and ignore switches during print moves.
I’m thinking the Z stage will be built properly perpendicular to the XY gantry, but with the bed mounted at like a 20 degree angle, maybe on a ball-mount to allow variable bed angle. Then Z probe bed tilt compensation (full rotation matrix style) prints the part in the right geometry.
Net result, a crooked frame of variable dimensions and crooked bed produce square prints of appropriate size. And it should be able to fold up quite a bit.
Each of the four XY crooked frame members will have an attachment point at the center of the beam. On the adjustable length beam(s), a self-centering double gear rack will provide the center point. Four beam centers produces a parallelogram. Then linear rods are run between attachment points to produce a parallel mount for a gantry. The gantry bridge rods extend past one carriage to accommodate the differing parallelogram width. Both X and Y run on screws which can also extend past the gantry. So now we have a working XY gantry mounted on an arbitrary-shaped frame.
Then mount non-contact switches at MID span for XY homing, set switches as 0,0, and ignore switches during print moves.
I’m thinking the Z stage will be built properly perpendicular to the XY gantry, but with the bed mounted at like a 20 degree angle, maybe on a ball-mount to allow variable bed angle. Then Z probe bed tilt compensation (full rotation matrix style) prints the part in the right geometry.
Net result, a crooked frame of variable dimensions and crooked bed produce square prints of appropriate size. And it should be able to fold up quite a bit.
Ryan Carlyle
Apr 8, 2019, 10:13:50 PM4/8/19
to 3DP Ideas
Midpoints of any quadrilateral form a parallelogram. Imagine segments AD and BC are telescoping with variable length. The gantry can collapse for easier shipping and then extend in an arbitrary position to print.
Reply all
Reply to author
Forward
0 new messages