Tuesday, March 25, 2014


I have been building up the second CRAMPS board to verify the low-cost PCBs from off-shore really work as expected.  Since I'm doing another build anyway, I have been documenting the process with pictures and have created a CRAMPS 1.0 page with assembly details, ECOs, and other board details on the RepRap Wiki.  Follow along there if you're interested in tracking progress or are building your own board.  If you are interested in building up a board and can hand-solder the six fine-pitch SMT parts, send me an email and I'll try to get you a bare PCB.

Meanwhile, work is progressing on the next version of the CRAMPS board thanks to Murry Lindeblom.  Murry has been doing some great with KiCAD getting the schematic whipped into shape while I've been busy with other projects.  The main changes are to simplify the design, reducing parts count and making the board cheaper to produce.  See the list of ECOs on the RepRap Wiki for hints about the next version.

Monday, March 24, 2014

tBone Cape

The tBone 3D printer / CNC cape has been announced for the BeagleBone.  It looks like an interesting board with some unique features.  I like the use of more sophisticated stepper driver chips from Trinamic, and the support for higher-current stepper motors.  This board seems to have an Arduino on-board to handle real-time, so it will likely take a bit of coding to get it working with MachineKit which is expecting to control real-time motion with the PRU.

For more details, see the tBone website and their Indiegogo campaign.

Wednesday, March 19, 2014

MRRF Report

by Charles Steinkuehler

I feel the Midwest RepRap Festival was a great success.  My primary goal was to increase awareness of using LinuxCNC and the BeagleBone to control a 3D printer.  I was interviewed by Hackaday, talked so long to many key people in the RepRap world that my throat was sore, and have been requested to write an article on using the BeagleBone for a major publication (stay tuned for details!).  It was great to meet and talk to Josef Prusa, see Nicholas Seward again, meet Drew Fustini and Jason Kridner in person, and talk to everyone else at the Festival.

From the many discussions I had with folks looking for something more than an AVR based Arduino to use for machine control, several key tasks became apparent.  Look for these to be implemented soon, and feel free to help out if you can!

  • Debian Packages:  With the BeagleBone soon to ship with Debian installed by default, it is important to have packages available so LinuxCNC can be easily installed without having to build from source.
  • Custom interfaces and/or an interface API.  No one likes current interfaces available with LinuxCNC for 3D printing (myself included).  There needs to be a simpler interface geared towards 3D printing, and many folks were interested in running headless or with a custom simplified interface (think small text LCD or OLED panel vs. Linux desktop).
  • Example configurations:  Ease of use is critical, and besides being put-off by the LinuxCNC interfaces, most people were further discouraged by the complexity of configuring a machine.  This needs to be made a lot simpler.  Part of the power of LinuxCNC is it's flexibility, but that makes for complex configuration files.  Most of this can be hidden for standard machines (Cartesian 3D printer, linear-delta printer, gantry desktop mill, etc), leaving users to simply adjust machine specific details like axis gain.
There are some other big news items and projects I learned about at the Festival, but can't talk about publicly yet.  Stay tuned, watch for news from the Beagle Board folks, and you might want to keep an eye on what Josef Prusa is up to.

Wednesday, March 12, 2014

Crunch Time (MidWest RepRap Festival)

by Charles Steinkuehler

At the last minute, I got permission from my wife to attend the Midwest RepRap Festival in Goshen, Indiana this weekend (March 14-16).  That means I am now in serious crunch mode (along with the rest of the folks working on MachineKit) to get some demos working for the event.  I am working on finishing my Wally beta build so I can get it printing using real-time kinematics powered by LinuxCNC.  Michael Haberler and Alexander Rossler are working on a networked HAL demo that will show some of what is possible using the new real-time foundation at the heart of MachineKit.  Everyone else is helping with marketing materials, website updates, and all the other tasks that need to get done.

If you're going to the festival, look me up.  In addition to demos, I plan on bringing what is likely the largest assortment of hardware CNC capes for the BeagleBone ever in one place for show and tell  I will have what is currently the worlds only working CRAMPS board, one of a very few K9 boards in existence, as well as several others.

Stay tuned for updates and pictures/video from the Festival!

Sunday, March 2, 2014

CRAMPS Board Working

I have the first CRAMPS board assembled and mostly checked out.  There are a few minor glitches (see the ToDo file on github), but so far no serious problems or anything that makes the current PCB design useless.  I have an example LinuxCNC configuration available if you want to try things out on your own.  The configuration requires you install my new Universal BeagleBone I/O Overlay, which can be used to switch between most "interesting" hardware options for the various I/O pins without having to load multiple overlays or possibly merge several incompatible overlay files by hand.

If you're a fan of scary movies, I even made a video:

Thursday, February 20, 2014

Linear Delta Printing with LinuxCNC

Reviewing my posts, it looks like I somehow missed sharing Brandon Heller's incredibly awesome video of his Kossel linear-delta printing parts for a GUS Simpson (a grounded delta).  I printed up a few sets of the GUS Parts on my MendelMax and will vouch for how hard they are to print.  I've gotten my Kossel based printer to move via LinuxCNC, but it's no where near as far along as Branon's.  Great work!

Friday, February 14, 2014

The King Is Dead

Long Live the King!

...which is a round-about way of saying this blog has a new address!


When linking to anything on this blog, please use the new blog.machinekit.io domain.  Of course any URLs you have stashed in bookmarks or web pages will still work thanks to the wonders of modern networking.  :-)

Over the next several weeks, look for more additions to the machinekit.io network family, including a main website and some more technical posts on how the new architecture differs from the existing LinuxCNC, what bits-and-pieces are available, and how to build useful machines that work in hard real-time without spending man-years writing code.

Lots more details coming soon, as well as lots of handy usable software.  I'll be releasing a major update to the PRU code which will support encoders (for full servo systems w/o an FPGA) and a consistent pin numbering system you won't have to be a programmer to understand.

What is MachineKit, I thought it was a BeagleBone uSD image?!?

What is MachineKit?  I like to think of it as a "Robot Army Construction Set", but something like "controlling movable things" is probably better.  Give us a bit to work up a good slogan...we're all engineering types here and marketing taglines are hard.  :-)

But yes, the work I've done to get LinuxCNC running on the BeagleBone does have a downloadable image named MachineKit.  However there is more to MachineKit than just an SD card image.  MachineKit is Michael Haberler's vision for LinuxCNC being worked on by a small (but very talented and busy) group of folks.  The idea is to take the "guts" of LinuxCNC back to it's modular roots and into the future by providing an assortment of pre-built, ready-to-use drivers, routines, and interfaces for machine control.  Think something sort-of like Lab View, but with hard-real time performance and the ability to (safely) control anything from room-sized CNC machines to desktop 3D printers to self-balancing robots and autonomous quad-copter drones (hmm...did I mention Robot Army Construction Set?).

Background and dry technical overview to help you sleep tonight

One of the first major features of the new MachineKit code was porting LinuxCNC to run on a variety of real-time kernels.  LinuxCNC used to support several real-time options but recently had degenerated to only working with RTAI.  With the introduction of several new RTOS options (including Xenomai and RT_PREEMPT), it became possible to run LinuxCNC on the BeagleBone.

Other enhancements being worked on include a completely new messaging scheme to communicate between various components of the LinuxCNC/MachineKit world.  This not only simplifies several existing use cases, but opens up many new exciting possibilities.  Real-time and user-interface components will be able to be distributed across multiple platforms (ie: different physical CPUs connected via networking, shared memory, or some other means).  The simple version of this is running the user interface on something like a tablet or desktop PC, while a BeagleBone or other embedded processor runs the real-time code.  But even more is possible.  The new scheme has been designed to allow multiple CPUs to participate in real-time operations, so you could potentially run floating-point real-time motion path generation on the ARM of the BeagleBone, use the PRUs as "first-class HAL citizens" (rather than the pseudo-hardware interface they are currently), and even something like an AVR or ARM CortexM to off-load various other bits of HAL code.

The vision of MachineKit is to return LinuxCNC to it's modular roots (you can theoretically swap out the motion planner, gcode processor, and other core parts of LinuxCNC, not that anyone does these days...it's code left over from the old times) and be able to more easily mix-and-match different pieces, potentially even having more than one component that performs the same or similar function.  As a simple example, perhaps you have a linear delta 3D printer (Rostock or mini-Kossel).  Usually, you want the motion controlled by the gcode processor for printing a part.  But sometimes you may want a complex program directly communicating with the motion planner to do things like non-trivial homing and auto-calibration (auto-probe a bunch of points on the bed and do a bit of linear algebra magic to determine the real physical characteristics of your machine, like arm length and tower offset, rather than trying to measure everything to sub-millimeter accuracy).  MachineKit intends to make this (and much more) not only possible, but easy to implement.