Here's a robot video that I found interesting:
Featured on http://www.robots-dreams.com
When I posted about 3-D printing without support before there were several questions about my printing setup, and some of the techniques that I use. Also, I've also changed some of the processes since then I thought it a be a good idea to post another video.
My current printer is a RepRap-based Tantillus with 100 mm x 100 mm x 100mm print envelope.
I print primarily using PLA, because it's more ecologically friendly and is food safe. The system is small enough to be portable, and allows me to print any place that's convenient, even on the dining room table. All my printing is done using the built-in SD card interface so that I don't have to worry about computer or communication problems, especially during extended prints.
I've switched from using the 3M blue tape and am now printing directly onto glass. I precoat the glass with a thin layer of water soluble white glue -- the type that is used in preschools and kindergartens so that it washes off easily. Although the glue layer thickness is a little tricky, is quite durable, and I've been able to print more than a dozen parts without having to recoat the glass.
I seldom experience any warping problems, but when it does occur it's always in the front left hand corner which leads me to suspect that my print bed isn't perfectly level.
Most of the parts on the printer were printed in PLA, including the gears. The herringbone gears that drive the X and Y axis shafts have effectively zero hysteresis which results in extremely good print accuracy and smooth surface finishes. Of course it helps that almost all of my parts are printed with hundred micron layers. I have gone down as far as 40 µm layers for parts when it was necessary.
The printer uses high-performance fishing line to move the printhead, which is always produced consistent results with no slipping or stretching.
Since it's a bowden tube design the printhead has low mass and can move quite fast.
This test part is a 30 mm hollow cube. The G code was created using Kisslicer, which produces consistently good results that are dimensionally accurate. I used to use Slic3r, but found it to be too unpredictable and often frustrating.
The primary challenge with this part is to bridge the gaps at the top of the cube without support. The printer, with a little bit of fine tuning, is definitely up to the challenge. There are a few stringers, but those cleanup easily with an X-Acto knife.
I really believe that a 3-D printer is a tool that should provide consistently reliable results with a minimum of operator intervention. So far, Tantillus fulfills those expectations better than other printers I have tested.