Micro-positioning stage.
Author: Matt Reilly
A simple sliding table that can be positioned via a 4-40 or M3 screw.
It requires a 3D printer and a Flavia stick. The stick is used for the linear bearing. Other things might suffice, but I had some of these around.
This is what a Flavia stick looks like:
This is a single-axis version of the micro positioning stage.
And here is an XY positioner.
The source for the components is in OpenSCAD (http://www.openscad.org). It is better than STL for distribution, as normal humans can read and write it.
I try to provide some commentary in the source code where things aren't obvious or where I expect you'd want to make some changes. Most of the good comments are in Common.scad where all the dimensions are defined.
The prototypes and test widgets were printed on a Monoprice Maker Select Plus using Silk PLA at 230C. The slide was printed flat side up, sitting on the four feet. This provides better dimensional control of the horizontal parts of the feet which must fit nicely inside the Flavia sticks.
I used Cura version 3.6 for the slicer. (Yes, I know they're on to newer versions, but 3.6 works and I don't have to worry about losing a lot of configuration data.) The infill density was 30% and the support density was 10%. YMMV and I didn't do an exhaustive study to find the perfect combination.
All dimensions are in inches, since that's what I have in the way of calipers. Yes, it is kinda 19th century, but I've got a 21st century computer in front of me that can multiply and divide by 25.4.
That said, the models here are all set up for M3-0.5 screws. That gets you 500 microns per turn of the setscrew. Skip the coffee tomorrow morning and you might get 10 micron resolution. If you are reading this in black and white, then you may only have access to imperial hardware. A 4-40 screw and nut will work just fine. You'll get about 25 mils per turn which ain't so bad for something you made in your basement.
Pretty much as in the picture. (See the picture. The background is the seat cushion on my lab chair. Sometimes I use red felt, but the shiny red plastic doesn't show up so well.) The design files here have added mounting holes at the corners of the base and slide. This allows you to bolt a Y axis base onto an X axis slide, giving you two dimensions to fiddle.
I've found that cyanoacrylate will hold the Flavia sticks to the PLA. I wouldn't hang a man in a hardhat from it, but the bond is serviceable.
This is not a laboratory grade device. It does work. (Or at least the single axis version works. The printer isn't done with the second set of parts yet.) The lash in the movement isn't going to worry the folks at Starrett any time soon. If you obsess about such things, turn the drive screw CCW to push the table away from the riser block. Then turn the screw head CW to draw the table toward the riser block. That will probably get you the smoothest motion.
If that isn't adequate, then a trip to Ebay will probably get you something that even the most discerning user would find suitable.
There is no displacement indicator -- an indicator is superfluous for my purposes. If it is important to you, I would encourage you to either
- Design a solution and try it out. When it works, post it to github.
- Take a look on Ebay where you will probably find something that even the most discerning user would find suitable.
The travel is limited to the length of the "leadscrew." The assortment I bought from the "High Precision Machine Parts and Screen Door Company of Guangzhou" topped out at 20mm. However, Grainger has graciously offered to supply me with 50mm screws for a modest charge. I suspect you could make similar arrangements.
Q: Why a normal screw and not a threaded rod?
A: A threaded rod would be ideal, but small diameter threaded rod is hard to find. Much of the stuff you find on Amazon and other places will be threaded at the ends. (I think the intended use is for servo linkages.) One could thread a rod with things found around the house, but life expectancies being what they are, I can't see it myself.
Q: Why is it so small?
A: I wanted this to be easy to print. This fits on every 3D printer that I know.
Q: How long will this take to print?
A: It all depends on your printer. I printed both pieces on my well-worn Monoprice Select Plus at 0.15mm resolution in a little over 4 hours.
Q: Why a pocket for the lead screw head, rather than a hole.
A: Materials, variations in printer quality, phase of moon, and a million other things conspire to make close tolerances hard to achieve with a typical home 3D printer. Since the hole in the base block would need to line up pretty closely with the hole in the slide, all of those variations would need to be controlled. Counting on that is just bad engineering.
When you hear someone say "the device is built to exacting tolerances: every part must be exactly the right size" you are either hearing
- someone who is talking through their pants
- a sales pitch for a luxury item whose cost and price appeal to a certain kind of person
- a description of a precision device that you will never be able to afford
- a story about crap engineering by people who are too lazy to think about reality and are willing to either endure lots of rejects at the end of the line, or are going to use you for "quality" control.
Good engineering will design within the capabilities of the technology and the properties of the material. I'm not a professional Mech-E, but I respect them enough to at least try.
Q: Will this work with ABS? PETG? Mother-of-pearl?
A: I haven't tried it. Someone should. (Well, maybe not with mother-of-pearl.)
Q: Why do this?
A: I've got a 3D printer. You've probably got a 3D printer. How many tiny plastic boats does one person need?
Q: How can I prepare for an exciting career in the world of mobile pet grooming?
A: I can't help you there.
See the LICENSE file for copyright and conditions for re-use.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent