Friday, April 26, 2013

The Fleet is in!

A friend of ours asked if we could print some battle ships his daughter needed for a diorama she was making as part of a school project - I said sure!  Then went hunting for models.  There are some great models in the Sketchup warehouse, but they are all crazy detailed and not printable.

Searching around Thingiverse turned up this model:


It looked like it would work and a few folks had printed it already, so that was selected.  In my opinion the hull is much too tall for the ship (maybe it was done this way so it would float).  I loaded the model in Blender and reduced the height.

Loaded it up in Kisslicer and produced the G code file that let me print these:
The left image was printed on the MM1.5+ and the right hand one was 3 ships at once on the MM2.0.


Below is a video showing some of the ships being printed -- this is from the MM2.0 with faster acceleration than we'd previously used.


Monday, April 22, 2013

Printing with ABS

After getting comfortable printing PLA, we decided to buy some ABS and nylon.  I opened up the ABS roll today and after printing a few test cubes, I loaded up a Spocktapus. Why Spocktapus again? First, I've printed it enough times to know what it's supposed to look like and what it's not supposed to look like. Second, it will annoy Colin (they're like Tribbles around here).

This was printed at 0.25mm layer height at 230C. It curls a lot more than PLA, which already curls too much and the smell is a lot stronger than PLA. Here's a video and final shot.



Not the best print ever, but I just wanted to crank this out quickly.

Friday, April 19, 2013

More worm work

My curiosity with worm gears continues...

The gears I printed yesterday were too small to make workable, and I've since modified the OpenSCAD file to give me better sized parts.

I need to make a mount that will hold the NEMA17 and the 8mm rod and a couple of bearings along with the gears.  This is what I have right now:

Like I said yesterday, OpenSCAD is a wonderful program (with limitations) and I'll use it to design and build the mount.  Hopefully it'll be done soon and I can see the worm drive in action!

Thursday, April 18, 2013

Worms, gears and OpenSCAD - Oh My!

I've been wanting to play around with worm gears and stepper motors for awhile, and today I had some time to experiment.  I downloaded this http://www.thingiverse.com/thing:8821 from Thingiverse today and after a bit of work, got it rendering correctly with OpenSCAD (http://en.wikibooks.org/wiki/OpenSCAD_User_Manual).

Initially I printed the gears the same size they were in the original file and was quite surprised by how well they turned out:



The fitted pretty well together and even turned as you'd expect.  The actual worm gear had some underhanging PLA under it's flanges, but it was better than I thought I'd get.

I then decided to reduce the size of them and try it again, this time with something I'd actually use on a stepper motor:

They came out beautifully - and don't they look cool!

When I've got them in a jig and hooked up to a stepper I'll post updated pics & video.



Giant Lego Vader

A friend of mine had asked me to print him a Giant Lego Darth Vader he'd seen on Thingiverse -- http://www.thingiverse.com/thing:50300 and I said I'd be happy to.

After buying some cheap black PLA from an EBay supplier and failing to get the thing to stick to the bed or even itself, we ended up ordering a lot of Ultimachine Black PLA (which we use for a lot of things now) and it prints really well.

I finally got around to finishing the print and figured I'd share just before I dropped it in the mail.




These are all the parts laid out on the desk - they're all printed in black, except for the light saber blade which is done in a translucent red.










This is a close up of the body - you can see the layer height is pretty uniform and the body detail looks great, even close up.















His legs were also printed at a 0.25 layer height - I'm pretty happy with the results.
















Finally the light-saber handle - this is printed vertically with a built-in support.  Kisslicer seems to generate some broken paths when slicing this file (around the base) so I printed it with a grid to get it to stick better to the bed.  When the print is finished the grid and the extra support are cut away leaving a hole for the blade to be inserted.  I think this was printed at 0.2mm layer height.


Wednesday, April 17, 2013

Winterfell on the MendelMax 2

The MM2 has been up and running for a few weeks now so I printed another Winterfell model to test the print quality. The print took about 3 hours at 0.1mm layer height, 1 loop, no infill at around 35mm/s.




Tuesday, April 16, 2013

Fun with Git

I've started collaborating on a software project with a friend and we're using a remote Git repository to share the code -- anyway, after banjaxxing some of his changes he suggested I read this:


It's a pretty good book and free -- well worth the read if you intend to use Git (especially with more than one person, and even more especially if that other person is even grumpier than me).

Setting up a Sanguinololu on a Mac

The printer kit we'd bought previously came with a Sanguinololu board as the controller.  This is basically a RAMPS device that has been squashed into one board with all the functions for controlling a printer built in.   You can read more about it here: http://reprap.org/wiki/Sanguinololu.


It's a pretty nice bit of kit and is certainly easier to wire up than your standard RAMPS, but is limited to one extruder, no support for fans and we're using it with a relay to turn the heated bed off and on.  From what I've read I think this is just the original kit designer being cautious.

It's also missing the extra LEDs present on some RAMPS which let you know when the heaters are on, and the expansion header is different so it won't work with the standard LCD screens or SD readers.

Anyway, it's what we have and I got it nicely wired up to drive our new printer.  When we setup the first one, we did so with Windows and there are plenty of instructions on how to do that.  This time I wanted to use my Mac to do the initial flash and setup.  Figured I'd share my experiences.

Now, I'm not saying everything I did is correct -- but it worked for me and hopefully it might work for you too...

First, I installed the FTDI drivers for the serial USB chip that's on the board.  I downloaded this from http://www.ftdichip.com/Drivers/VCP.htm as directed by the standard Arduino website (http://arduino.cc/en/Guide/MacOSX).

On the FTDI page there are a few options - for me (OSX 10.8.3) I grabbed the 64-bit Intel file and mounted the disk image once it was downloaded.

From the disk image, I then installed the latest version which seemed to work fine for me.

Now I had the driver installed, I next downloaded the Arduino software for Mac OSX from http://arduino.cc/en/Main/Software.  The Mac version is provided as an App and you can just execute the downloaded file directly.  I did copy the App into my Applications folder so I could find it later, and it would show up on LaunchPad.

We're used to using Marlin, so I downloaded the latest version of the Marlin firmware from https://github.com/ErikZalm/Marlin -- just download the latest release, unzip it and put it some where you can find it later.

So, before you can build the firmware and send it to your Sanguinololu you need to modify the Arduino tools so they know about your board.  This is pretty straightforward, but neccessary.

From http://code.google.com/p/sanguino/downloads/list download the latest release and unzip it.  Now go back to the Arduino App you put in your Applications folder and right click it.  Select 'Show Package Contents' and left-click.

This will open a Finder window - double click the Contents folder, Resources folder, Java folder and finally the Hardware folder.

Grab the sanguino folder you downloaded and unzipped earlier (from http://code.google.com/p/sanguino/downloads/list) and drag it into the Hardware folder.  Your hardware folder should now look like:

With that done, you can close the Finder window and run the Arduino App from either LaunchPad or from your Applications folder.  If you installed the Sanguino folder correctly, you should see the Sanguino options at the bottom of the board list:

If the Sanguino options are not there, then you did something wrong -- maybe you already had the Arduino software running when you added the Sanguino extensions - if so quit it completely and re-launch it.

Now you can load the Marlin firmware and configure it for your Sanguinololu.  Navigate to the folder you had downloaded and extracted the firmware you downloaded earlier.  This should be a folder called Marlin-Marlin_v1. Open the folder, then open another folder called Marlin.  There should be a file called Marlin.pde.  That's the project file the Arduino software opens.  You can right click on the Marlin.pde and tell Finder to open it the Arduino App.  Once loaded, save the file and the Arduino App will let you know that the project extension has changed.  Go ahead and let it save a new copy of the project file as Marlin.ino.  In future, you can just double click Marlin.ino to load the project.

There are just a few more steps to get this working on the Mac -- you need to select your board type from the Tools/Board menu - I have a Sanguino ATmega1284p 16Mhz, so that's what I selected - you'll have to pick the entry to match your board.

You need to select the serial port that your board is on - this will only appear if the board is connected to the Mac and is powered up (USB power will do for this).  In my case I use the following setting:
dev/tty.usbserial-A104M4B

Yours might be different, but it should have tty. at the start of it. You can try compiling the firmware and sending it to your board to make sure the connection works, but before you do that, make sure you set the board type to 62 in configuration.h.

That's it - you now should have your Mac talking with your Sanguinololu and you can configure the firmware to suit your printer.

Enjoy!













Friday, April 12, 2013

PVA Glue

After my hairspray fun yesterday, good old PVA glue came to the rescue.  Makes the print bed look like crap, but things certainly stick well to it!


Thursday, April 11, 2013

Printing without hold

I needed to print a part today for a new printer design we're working on -- more about that later -- the MM1.5 was busy, so it was time to try using the MM2 to print it.

We've been printing on the MM2 pretty well, but had kapton tape layered on the glass.  Unfortunately due to an earlier accident (someone typed 1.2 instead of .12 as the Z position in a G92 command) the bed ended up looking like this:

As you can see, the head ended up scratching up all the tape as it moved back across the bed.  Fortunately it didn't scratch the glass or even break it.

So the tape was removed, the bed heated and printing fun began.














First layer seemed to go down ok, but after a few layers I felt comfortable enough to walk away and let it continue its work.

Fortunately, I came back and caught it doing this before I stopped the print.  The print had come free of the bed during printing and was being dragged around by the head as it spewed pumpkin colored PLA all over the place.

Not good.








Based on a recommendation from our  resident 3D printing expert, I decided to give the glass a quick blast of Garnier Fructis Style -- with EXTREME control -- initially the  results seemed to be positive, the printed part stuck nicely to the bed and everything seemed to be working out just fine.




















Sadly to say, this was not the case. As the print approached the end of its run, the part came loose and I ended up with a melted mess on top.
I think I'll be able to use the part at least to test the belt motion it was designed to fit as the mounting holes are mostly complete on both sides.

I guess we'll end up covering the bed in Kapton tape again.

On the plus side, the orange PLA looks great against the blue plate on the MM2 - if only it had stuck better.