Those of you who got a chance to stop by Maker Faire this weekend (or read our earlier blog post) are probably wondering when you too can experiment with dual extrusion heads. The answer is: now! We’ve integrated tools into ReplicatorG that simplify the process of creating a two-material print. They’re not in the released version of ReplicatorG yet, but they are available in a git branch for experimenters who want to get a jump start on the future. And who doesn’t want to get a jump start on the future? Details for the bold below!

First of all, you’ll need to check out the “dualstrusion” branch of ReplicatorG. As always, all the code is available from our github repository. If you’re new to running ReplicatorG from source, take a look at the build instructions. Once you’ve cloned the repository and successfully run ReplicatorG, you’ll just need to check out the experimental branch and rebuild:

> git checkout dualstrusion

> ant run

Once you’re up and running with the experimental code, read Ben Rockhold’s guide to designing for dualstrusion, which will explain how to design and print objects for dual extrusion.

Additionally, if you’re using multiple toolheads and having trouble with communications, you’ll want to try building the latest firmware from source, as the current version in git implements a protocol change that greatly improves operation for multiple toolheads.

A special thank you goes out to Ben Rockhold, Noah Levy, and Will Langford, who made all this awesomeness happen.

Good luck, and as always drop me a line or a comment and let me know how it goes!

Bad Y-connector on the left; good Y-connector on the right

We’ve discovered that one of the parts we’ve shipped with some recent Thing-O-Matic kits, the Molex Y-Splitter cable, isn’t up to spec. The copper wires in some of these connectors are thinner than expected, and has trouble carrying enough current to the power-hungry extruder controller. This can cause the cable to grow hot, or even burn out. Luckily, there’s a simple work-around for this problem.

How do I know if I have a bad splitter?

The bad parts all have bright white plastic shrouds around each connector. If you look at the 4-pin Molex connectors that are attached to your power supply, you’ll notice that the plastic is semi-translucent. If the connectors on your Y-splitter are noticeably Whiter and Brighter™ than the ones on your power supply, and the cables say “20 AWG” in tiny characters on the side, you probably have one of the problematic ones.

Bright white (left) == lame, translucent (right) == sweet

What do I do if I have a bad splitter?

Don’t panic! There’s a workaround. Although the under-spec Y-splitter can’t handle the power delivered to the extruder controller, it can handle the power required by the stepper controllers without too much trouble. All you need to do is unplug and rearrange the power connectors inside your bot:

1. Turn off the power to your bot and unplug if from the wall and USB.
2. Open the bottom of your bot.  You’ll probably need to disconnect a few of the wires, particularly the endstop cables, to open it all the way.
3. Disconnect all the 4-pin Molex connectors from the stepper drivers and extruder controllers.
4. Plug one of the 4-pin Molex connectors that come straight from the power supply to power your extruder controller.
5. Use the Y-splitter(s) to hook up your stepper controllers.

Please note that if you have multiple Y-splitters, you shouldn’t chain them– hook one up to each power connector coming from your PSU.

That’s it! You should now have a much happier bot.

We’re still in the process of diagnosing this problem, so if you’ve got some input (or you’ve already burned out your Y-splitter) please drop us a line and we’ll get in touch as soon as we can.

Thing-O-Matic on top, Cupcake below.

This release fixes the too-hot Thing-O-Matic profiles mentioned in yesterday’s post.  It also includes example STLs and convenience gcode scripts accessible from the “File” menu.  (See the ‘Attribution.txt’ file in the ReplicatorG/examples directory for more information about the sources of the examples.  Hint: they’re all from Thingiverse!)

As always, let us know what goes right and what goes wrong.  Changelog below!

0023 ReplicatorG
* Bugfix release
* Fix machine profiles to handle correct endstops and platform sizes
* Fix too-hot 0006 TOM profiles
* Added emergency stop button to control panel
* Readded raft checkbox on Skeinforge 0006
* Removed Skeinforge 31

There will be a new release of ReplicatorG tomorrow afternoon that corrects this issue.  Until then, use the Skeinforge 35 if you’re generating gcode for Thing-O-Matics, or you may end up with a sad extruder, like a certain Mr. Pax:

ReplicatorG 0022 is out, and… I don’t even know where to start! There’s so much goodness in this release that it’s just absolutely crazy. Thanks to Marius Kintel, we now have Skeinforge 35 integrated, with the ability to generate support material for builds. Erik DeBruijn has improved the feedback from Skeinforge, and added code to help ReplicatorG work well with other RepRap-derived machines. We’ve added profiles for the Thingomatic and drivers for Generation 4 electronics. There’s new firmware for bots both old and new that will help improve reliability and performance. There’s new user preferences! Experimental features! Additional checkboxes! Fancy dropdowns! It’s just madness! Download it now!

This also marks the first ReplicatorG release where the lion’s share of the work was done by contributors. A bit thank you to everyone who pitched in over the past month and made this the best release of ReplicatorG yet by far. Changelog is after the jump!

Update: The Mac build has been updated to work with older 1.5 JVMs.  If you’ve had problems running ReplicatorG on older Macs, pull down the new version.  Thanks to Darren Bleuel for identifying the problem.

Update 2: Got Rob Giseburt’s name wrong.  Sorry, Rob!

0022 ReplicatorG

• UI for selecting basic skeinforge options
• Fix: Numerous fixups to connection, disconnection, initialization, post-upload and post-simulation behavior
• Fix: update firmware list immediately on URL change
• Optional icons for boards in uploader
• Machines.xml banished; machine descriptors are now broken up among files in the repg/machines and ~/.replicatorg/machines directories
• Preference panel ui to control logging level
• Experimental: multiple extruder support
• justjoheinz: added profile delete button
• nazdravi: check for “python2″ for compatibility on systems which default to python 3.x
• Erik: added 5D driver for RepRap’s and compatible firmwares (Ultimaker)
• Erik: improved feedback during skeining operations
• Erik: STL preview now is the actual build volumes configured in machines.xml
• Charles: Support for G10 and friends
• Marius: Added generalized skeinforge merge patch
• Marius: Added skeinforge 35
• Marius: Added build platform belt checkbox to control panel
• Marius: Added support for wait-for-platform
• Nick Starno: paraminator script
• Robert Giseburt: several integration fixes for OS X
• Marius, Charles, Nick, Rob Gilson, and others: lots of hard skeinforge configuration work!

Note: Want to print out the serious head pictured above? Find a serious head over on Thingiverse, by yzorg.

We’re looking for a programmer to help out with our firmware and software.  This is a full-time position at the botcave in scenic Boerum Hill, Brooklyn, just steps away from Atlantic Terminal.  You’ll helping to maintain the firmware for our current boards, as well as developing code for the latest and greatest.  You’ll also be working a bit on ReplicatorG, supporting new products, and coordinating with open source hackers around the world.

Interested?  Check out the job listing, and get in touch.  We’re looking forward to hearing from you!

This is primarily a bugfix release.  There were some situations in which stopping a build could put the communications between the bot and the host out of sync.  If you were sometimes experiencing “temperature inversion” in the control panel, where the extruder temperature and the platform temperature were reversed, this should solve your problem.  It should also fix cases where the extruder would get out of sync and stop extruding after an aborted build.  If you’ve been seeing a delay when you hit the pause or stop buttons (for example, the bot finishing the segment it’s on before stopping or pausing), that should be corrected too.

There are a couple of other small goodies in there too: exiting the app will no longer halt an SD card build.  There’s a first pass at a thingomatic profile for skeinforge 31, and the menus have been cleaned up a bit (thanks, charlespax and justjoheinz!).

As always, grab it from the download page.  Give it a whirl!

(One last note– there have been a lot of people contributing to ReplicatorG lately; so many that I haven’t had time to integrate most of the patches yet!  0022 is going to be the big merge, so if you’ve been hacking away on something, thanks in advance, and it should be in the main tree soon!)

This is the first release to feature multiple slicing engines– in this case two separate editions of Skeinforge.  By default, ReplicatorG will still use the version of Skeinforge we’ve had integrated since 0006, but if you want to play with the latest and greatest, we’ve also integrated the most recent Skeinforge release for your slicing and toolpath-generating pleasure.  Just select the slicing engine you want to use from the “gcode” menu, like so:

You can create and edit profiles for the new Skeinforge just as you would for the old version.  (Well, almost; we’ve added the create and edit buttons directly to the profile selection dialog.)

You’ll also notice that you can now launch either the extruder or the motherboard onboard preferences dialog directly from the machine menu, rather than having to open the motherboard preferences and then click on the extruder preferences button.  Which is handy, because we’ve got a few new extruder preferences for you to play with in the new v2.5 extruder firmware!

The new extruder firmware lets you remap mosfet channels and switch motor controllers, so if you’ve managed to burn out a chip or two, you can just change your configuration and keep going.

You’ll also notice that we’ve removed the automatic backoff feature from v2.5; we made this change to reduce stress on the motors.  If you miss this functionality, we recommend investigating the oozebane module in Skeinforge.

The v2.3 firmware for the motherboard doesn’t include any new features, but does increase the retry count on the extruder communications, which should help if you’ve been having trouble with longer builds.

As always, ReplicatorG doesn’t happen without you!  We have a bunch of contributors for this release:

• chlunde: Edge mode on ‘e’ keypress re-enabled
• Miles Lightwood, Eberhard Rensch, koenkooi: created/backported raftless to skeinforge
• Charles Pax: new profiles, machines.xml entry for thingomatic
• justjoheinz: about box fix

Apologies to anyone I’ve left out; drop me a line or let me know in the comments if I missed you.

Thanks, and have fun!