Hey! This thing is still a Work in Progress.
Files, instructions, and other stuff might change!
Rhombot
46
Likes
1806
Downloads
933
Views
Published on January 3, 2011
This thing was Featured on January 5, 2011
Description
Rhombot 0.3 alpha
See an animation of it on YouTube: youtube.com/watch?v=zSeg-Th1tRw
The Rhombot is a printable XYZ stage based on 3 chains of rhombus shapes. It is currently just an early alpha version. It's a rough sketch of the idea. See one of the two .MOV videos for how it is intended to work.
The Rhombot can be used with Makerbot motors, electronics, and extruders. All parts can be built in a Makerbot, or in itself. The Rhombot has no rods, belts, pulleys, nuts, threads, screws (other than the screws on the motor), sarrus hinges, moving towers, or moving motors.
That last point is important. The XYZ motors and extruder motor are all fixed. The X motor does NOT move a carriage carying the Y motor. The Z motor does NOT raise a platform carrying the extruder motor. Instead, all 4 motors can be permanently mounted. This means it will work with motors that are arbitrarily large and heavy. That should help when we start printing motors, since the early ones are likely to be large and heavy.
The green build platform has a Z coordinate that is the sine of the rotation angle of the NEMA17 stepper motor under the light blue panel. X is the sine of the dark blue arm's motor, and Y is the sine of the purple arm's motor. To use this to print, it would be necessary to modify Skeinforge to take the extra sine into account.
The build volume is large enough to print all its own parts. The animation even shows it building 4 of its own parts, as well as building a cube large enough to hold any of its own parts. The build volume actually extends beyond this cube, but is an odd shape. The cube in the video is the largest cube that fits within the build volume, oriented that way.
This has now been printed on a ZCorporation ZPrinter 650. It glues powder together to give an object similar to sandstone. This means the pieces are extremely rigid, but break easily. The printed Rhombot has hinges that move smoothly, with almost no slop at all in the direction of the hinge, nor orthogonal to it. The smooth movement is surprising, since the surface feels like sandpaper.
If you hold two panels that should stay parallel and try to twist one, it seems to allow almost no twisting at all. The small amount of twisting that it does allow seems to be from some of the hinge pins that are too thin. The hinge pins are wooden toothpicks, and don't have uniform diameter. They will soon be replaced with segments of 2.4 mm diameter aluminum rod.
The next step is to build this in plastic. It will be far stronger, but it may start having problems with panels bending and twisting. That's the big question. There are also several design changes that should make it into the next version.
See an animation of it on YouTube: youtube.com/watch?v=zSeg-Th1tRw
The Rhombot is a printable XYZ stage based on 3 chains of rhombus shapes. It is currently just an early alpha version. It's a rough sketch of the idea. See one of the two .MOV videos for how it is intended to work.
The Rhombot can be used with Makerbot motors, electronics, and extruders. All parts can be built in a Makerbot, or in itself. The Rhombot has no rods, belts, pulleys, nuts, threads, screws (other than the screws on the motor), sarrus hinges, moving towers, or moving motors.
That last point is important. The XYZ motors and extruder motor are all fixed. The X motor does NOT move a carriage carying the Y motor. The Z motor does NOT raise a platform carrying the extruder motor. Instead, all 4 motors can be permanently mounted. This means it will work with motors that are arbitrarily large and heavy. That should help when we start printing motors, since the early ones are likely to be large and heavy.
The green build platform has a Z coordinate that is the sine of the rotation angle of the NEMA17 stepper motor under the light blue panel. X is the sine of the dark blue arm's motor, and Y is the sine of the purple arm's motor. To use this to print, it would be necessary to modify Skeinforge to take the extra sine into account.
The build volume is large enough to print all its own parts. The animation even shows it building 4 of its own parts, as well as building a cube large enough to hold any of its own parts. The build volume actually extends beyond this cube, but is an odd shape. The cube in the video is the largest cube that fits within the build volume, oriented that way.
This has now been printed on a ZCorporation ZPrinter 650. It glues powder together to give an object similar to sandstone. This means the pieces are extremely rigid, but break easily. The printed Rhombot has hinges that move smoothly, with almost no slop at all in the direction of the hinge, nor orthogonal to it. The smooth movement is surprising, since the surface feels like sandpaper.
If you hold two panels that should stay parallel and try to twist one, it seems to allow almost no twisting at all. The small amount of twisting that it does allow seems to be from some of the hinge pins that are too thin. The hinge pins are wooden toothpicks, and don't have uniform diameter. They will soon be replaced with segments of 2.4 mm diameter aluminum rod.
The next step is to build this in plastic. It will be far stronger, but it may start having problems with panels bending and twisting. That's the big question. There are also several design changes that should make it into the next version.
Instructions
Print everything in the stl_to_print.zip file. Each filename tells how many to print. Connect them with hinge pins that are 3 mm in diameter.
The pieces actually have the build directions engraved on them. Two edges connect to form a hinge if they have the same number on that edge. For positive numbers, they should face the same direction, as if the hinge was formed while both panels were sitting on a table face up, with the writing visible on both. The numbers are negative if they face opposite directions, as if the hinge was formed while they were lying on a table with one face up and the other face down, hiding its writing. If a hinge has two sets of holes (so it can connect to two other panels), then there are two numbers given, with the first number referring to the highest hole (the one farthest away).
The pieces actually have the build directions engraved on them. Two edges connect to form a hinge if they have the same number on that edge. For positive numbers, they should face the same direction, as if the hinge was formed while both panels were sitting on a table face up, with the writing visible on both. The numbers are negative if they face opposite directions, as if the hinge was formed while they were lying on a table with one face up and the other face down, hiding its writing. If a hinge has two sets of holes (so it can connect to two other panels), then there are two numbers given, with the first number referring to the highest hole (the one farthest away).
You must be logged in to post a comment.
Webca
on
January 8, 2011
said:
When do you think you will try printing this out? Cant wait to see something like this in motion;)
Anonymous
on
January 5, 2011
said:
Here is a rushed movie from the frames provided. I discovered after the movie was uploaded that the frame sequence is in reverse, but you should get the overall idea.
merlinjim
on
January 4, 2011
said:
I love the idea of a low-vitamin printer. This might be very amenable to automated ejection; where is the travel area of the stage? Maybe you could do some more renders with the area as a transparent box?
tbuser
on
January 4, 2011
said:
The videos you uploaded don't work, I even tried uploading it to youtube and it couldn't process it either.
Collections
Things to Make
by 1125lbs
Open Source Beam and Rail System Parts
by SirGeekALot
Things to Make
by astrobiologist
License
More from LeemonBaird
This is very innovative. I like your idea of having non-moving motors.
The Rhombot does share a weakness with the Sarrus: the small scale stiffness depends on if there is any sideways stiffness with the hinges. That is to say if you push and pull one side of the hinge in line with the hinge pin axis, is there any slop? If so, this will appear as backlash in your m
ovement. This could be improved by giving the hinges some 'pinch'.
If the hinges have any slop, it will show up as backlash in the movement of the platform. Unfortunately, there are many hinges in series so the backlash will add up.
It is interesting to note that this should not be a problem wit
h the Z axis, since the weight of the platform will push all the hinges to one side. There may be a clever way of mounting the other axes at a slant so that gravity will preload them as well.
Yes, that's one of the issues I'm looking forward to exploring when this is built. I'm also curious whether the hinges will have slop in the other direction (as if the pin were too small for the hole), and whether the plates themselves will twist.
Mounting the axes on a slant would interfere with the motors being independent. I like that currently, each motor purely controls a single dimension. I'd like to avoid have the X position controlled by one function of multiple motors, and the Y position being controlled by a different function of those same motors.
There may be a simpler way to avoid slop motion in the direction parallel to the hinge pins. If there's a spring pushing one panel one way and the other the other way, that could do what gravity does for the Z plates. The easiest way to test it is with a rubber band, wrapped slightly diagonally across two plates. A better long-term solution might be a thin plate of plastic reaching out from one panel that pushes on the other.
We need to build it first, and see which of the many potential problems is the worst, and then work on fixing that one first.