Posts Tagged ‘cnc’
Inventables donated a ShapeOko 2 CNC Router “The Works” kit to Pumping Station:One, and we’ve been doing a group build over four sessions so far. It’s been a chance for people to learn about open hardware and CNC firsthand from the ground up, and participate in making PS:One’s next machine. We’ve had participants with various levels of experience working together, and I think everyone learned something new.
During the first session, we assembled wheels and bearings, and attached them to plates for the x, y,and z axes. We started the next session with parts that looked like this:
The instructions describe the z-axis assembly as the most intricate, and they’re not kidding. It took a lot of fiddly work and some mistakes to get here:
Thank you to everyone who helped so far! Our next build session will be Thursday, August 28th at 7pm in PS:One’s shop. We’ll be testing the motors, troubleshooting, and sending the “hello world” job. Come see the machine move!
Thank you Ron (I think) for the Session 2 photos and Allen for the Session 3 photo.
The adventure continues! We had a great turn out at the last ShapeOko build event. Now it’s time to assemble the gantries and do some wiring. Join us this Wednesday July 30 from 7-10PM in PS:One’s shop to see the machine really take shape, and maybe we’ll get to see it move, too. Learn about open hardware and the ShapeOko 3D carving machine. This event is open to the public and is great for newbies and experienced CNC’ers, too.
The sound board of the harp had always been the wild card. Elizabeth and I began the project last year knowing that we’d want to use a laser cutter to etch it with some kind of Victorianesque steampunk design involving gears. And we knew that PS:One’s Epilog, with its 24″ x 12″ bed and no feedthrough capability, simply couldn’t fit a 49″ sound board. While design focused on the brass panels, as we did have access to a CNC router, the sound board was left for later, especially because we didn’t know if we could find a smaller laser cutter with feedthrough, meaning we needed a design that could be etched in pieces, or a large laser cutter that could engrave the entire sound board at once.
Elizabeth and I were planning a trip to Madison, WI, and we heard rumors that Sector67 had a colossal Chinese import laser cutter. I reached out to them asking if they would be willing to help with this project, and Chris Meyer, director of Sector 67, responded, inviting us to their space. Knowing what we had to work with, Elizabeth was able to create the design in Adobe Illustrator. Read the rest of this entry →
For the past year, Elizabeth and I have been collaborating on a project using the amazingly cool CNC tools at Pumping Station: One. The goal: to build a harp. Not just any harp, mind you. A steampunk harp! The idea was to start with a kit (the Voyageur harp from Music Makers, 33 strings, cherry) but heavily customize it as follows:
- CNC cut brass panel inserts, inlaid in pockets routed in the sides of the harp
- The brass panels would be etched using a galvanic etching process, similar to the one used by the Steampunk Workshop to create their clockwork guitar. Elizabeth would design the shape and custom artwork (gears, of course!) for this.
- The sound board would be laser engraved with some type of steampunk design. The design is in progress, and we are searching for a laser engraver large enough to handle the sound board.
- Although not strictly steampunk, I’d considered adding RGB addressable LED lighting under the neck of the harp, which could illuminate the strings, as well as respond to the pitch of the strings being played.
First, we had to start by routing the brass, using PS:One’s CNC 3020 router. Elizabeth drew the design, including the brass outline and the pattern we will use when we etch the brass. We did this last spring at PS:One, and we ran into massive problems with the brass vibrating and breaking end mills. The project remained dormant for many months until a breakthrough: What if we glued the brass to a scrap board? That would at least keep it immobile so it couldn’t chatter and bind on the end mill. The good news: This worked! Success! The bad news…. The Go To Home button on Mach3 does not, by default, raise the spindle before moving it. And a clamp was in the way. The result: A badly bent spindle. Well, all was not lost….
Read the rest of this entry →
Our shop has a few Black & Decker drills. They’re not contractor grade tools by any means. I was using one the other day for a project, and as I was walking back to the tool crib to put stuff away I was idly clicking the drive direction switch back and forth, as one does with a drill. I heard a snapping noise and the button now moved freely, no longer engaging the electrical switch responsible for direction reversal. Bummer. I figured I’d pop it open and see if I could repair it.
The issue was a small plastic pin that engaged a switch with a matching cutout. Not a very complex mechanism. I drew the part up in Sketchup:
I skipped the nicely radiused leading edge, but this part is otherwise dimensionally similar to the factory piece. My initial attempts to use an entirely 3d-printed part failed, as the ~3mm pin was just not large enough to get a sturdy printed feature. I decided to drill it out and use a #2 screw to replace the pin. This one should outlast the rest of the drill.
At this point you might be wondering why I didn’t do that with the original part to begin with, and that’s valid. My only answer is, “Because I didn’t think of it.”
It’s a simple result, but it’s the sort of thing I love about 3D printing.
We’re throwing a 3D Printing Cage Match Party at 7PM at Chicago’s Beauty Bar.
See factory teams, local businesses, and hobbyists compete to print a medieval weapon as fast as possible, and then fire marshmallows at their slower competitors!
Our own DJ Adam Dzak will be spinning, Jim Burke will provide commentary, and PS:One’s Lulzbot AO-101 and Makerbot Replicator will both be competing as well.
Want to attend? RSVP (free) here.
Want to compete? Register (free) here.
More information at the 3D Printer Cage Match Homepage.
Please note that this event will serve as the CNC Build Club meeting on the same evening.
For the past few months, my CNC Build Club project has been building a <a href=”http://www.shapeoko.com/”>ShapeOko CNC mill</a> from a kit Jeff donated to PS:One to replace the machine hacked into a pick-and-place. The ShapeOko belongs to PS:One and will (hopefully) be a permanent part of the space.
On July 21st, I moved the machine to its home in the shop, finished wiring it up, and tested the motion of the stepper motors. It moved like it was supposed to on the x, y, and z axes, so I moved on to drawing the Hello World job (the ShapeOko logo) in the air. That worked perfectly, too. So now it was the moment of truth – time to find a drawing implement, tape it to the gantry, and send the GCode to draw the logo on paper.
After several attempts at finding the right pen or marker, and figuring out how to tape it securely, this was the result:
A very happy me, and a successful Hello World.
I started this project to learn more about CNC projects from the ground up. Along the way, I learned a bit about tapping, soldering, and Arduinos too. Here’s a look back at it:
This was the beginning:
Then I added the Z-axis:
Edward Ford, the Shapeoko’s inventor, happened to be at the space the night I finished the mechanical build of the Shapeoko:
After tweaking and tightening up the mechanical build, I assembled all the electronics I would need, mostly from donations to the project. (Thank you!) Edward came back for ShapeOko night as part of CNC Build Club, and we got the machine wired up. Unfortunately, the x-axis didn’t move properly, probably because the GRBLshield controller got damaged during rework. So Bart donated another GRBLshield, and Ryan did some heroic rework on its connectors, and this one worked!
Colin donated a Dewalt spindle to the project, and the next step is to get some end mills, test the machine’s milling, and certify some people. I’m also looking at installing some limit switches on the machine. Of course, there are also options like a different spindle, a dual-driven y-axis, or a more robust z-axis. Those will be things for the CNC Build Club and other interested members to decide on.
I got this far with more than “a little help from my friends”. Thank you to the people who offered help, parts, or advice (in no particular order): Jeff, Jay, Steve, Colin, Ryan, Edward, Cat, Bart, Jeremy, Fernando, Jesse, and Everett. If I accidentally left you out, I’m sorry!
We have a very special event planned for the next CNC Build Club meeting on Thursday, July 18th. We are going to do a Gonzo Build with the Make With Moto crew. Make With Moto is a Velcro covered Sprinter van filled with making equipment that is touring the country. In addition to the making equipment they have a bunch of unlocked and hacked Android smartphones. They have been doing hack-a-thons around the country.
We thought it would be awesome to do something that combines our CNC hacking skills with their phone hacking skills. We would temporarily remove the “a-thon” and do a one night Gonzo build.
The plan is to do a digitally controlled camera slider. The phone would replace the camera and control the motion. We are still seeing what is feasible on the phone control options, but ideally you could control the phone from a second phone and view live video or snap shots from the slider phone.
It will probably look similar to this camera slider that was done in the past, but it will have a mount for a camera and have all the drivers on the slider.
Inventables will be supplying all the CNC materials. As the details get more firmed up, I will edit this post so check back later. If you have any other ideas, let me (Bart) know. We may be able to get materials for multiple versions.
The project was to build a “single axis” drawing machine in one night. This machine uses two independently controlled carriages on a single piece of MakerSlide rail to control a pen at the tip of two linkages. The primary purpose of the project was a fun group build and a learning exercise in setting up a non-Cartesian machine using inverse kinematics. Kinematics in this case means mathematically describing the machine to the CAM controller. The One Axis DrawBot is a very simple non Cartesian machine.
The equations used are shown above. The ends of the linkages on the carriages end are at joint and joint. The pen is at pos->tran.x and pos->tran.y. The first two equations convert the desired pen location back to actual machine locations. They were plugged into the CAM program. The last two equations do the opposite and convert machine locations to the pen location.
We had one team assemble the machine, one team wire the electronics and one team setup the controller. It took about two hours to complete that phase. We try to use newbies wherever possible, so adding solder training into the mix usually adds a little time. The next step was to setup the CAM controller.
We borrowed the CNC router computer and control box to run the machine. This has Mach3 CAM controller software on it. Mach3 has a “formulas” feature that we used to enter the kinematics. We quickly had the machine running, but it was soon clear that Mach3 was not completely up to the task. In the formulas mode, it appears to disable the DROs (digital read outs) which tell you exactly where the machine is. It was also difficult to home or tell the machine the current location. Moves in the Y axis are non linear and need to know the current location. This resulted in Y axis moves that were not 100% accurate.
The other problem was coordination. If you tell a CNC machine to move from X0, Y0 to X1, Y0, it accelerates up to the desired speed then decelerates to the end point. If you tell it to move from X0,Y0 to X1 Y10, the two axes are moving different distances, so it needs to coordinate the different axes speeds and accelerations. The X axis would move quite a bit slower to coordinate with the longer Y distance to get a straight line. Mach3 was coordinating the two machine axes, but it was not coordinating the pen axes. Moves in only X or only in Y were nice and straight, but moves in both X and Y had a bit of a curve to them, but they did accurately arrive at the end point. A graphic with a lot of short moves would not show any on the problems above so we ran a quick “PS:One” graphic. Watch the video and be sure to wait for the applause.
We will switch to using EMC2 (LinuxCNC). This has a true inverse kinematics feature that should fix the problems.
There is some discussion on the EMC user mail list to help us with this project.
Update 6/23/2013: Here is the latest (untested) version of our LinuxCNC kinematics file mykins.c