Prize Winning Mostly 3D Printed CNC Mill (just download STLs)

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Winston

Lorenzo von Matterhorn
Joined
Jan 31, 2009
Messages
9,560
Reaction score
1,748
Just found this:

MPCNC

The MPCNC runs on inexpensive 3/4" ID (23.5mm OD) or the International Edition version (25mm OD) galvanized steel conduit. In the US it is under $4 per 10', with just under 20' needed.

US edition - https://www.thingiverse.com/thing:724999
International edition - https://www.thingiverse.com/thing:790533

Ryan Zellars shares designs for impressive mostly 3D printed RepRap CNC machine

https://www.3ders.org/articles/2015...ive-mostly-3d-printed-reprap-cnc-machine.html

MPCNC web site:

https://www.vicious1.com/

Parts list:

https://www.vicious1.com/blog/parts/

By looking aorund his web site, I also found a link to this freeware:

DXF2GCODE: a tool for converting 2D (dxf, pdf, ps) drawings to CNC machine compatible G-Code

https://sourceforge.net/projects/dxf2gcode/

[video=youtube;qJfYTv88YvI]https://www.youtube.com/watch?v=qJfYTv88YvI[/video]

With <$100 2.8W laser:

[video=youtube;WM2AMtxMhMk]https://www.youtube.com/watch?v=WM2AMtxMhMk[/video]

Other videos:

https://www.vicious1.com/videos/

Mid5.jpg
 
I embarked on starting to print the parts for this a few days ago. By my calculation, with my speed and infill settings, it will take over 150 hours of 3D printer time to make just the basic parts. Probably another 30 to 40 hours for the add-ons and options. 2 full kilograms of filament consumed by the time I am done. I plan on making most of the parts in PLA, but a few in ABS.

Exploring Thingiverse, there are dozens, if not over a hundred different options or add-ons. I plan to use the full size Rambo 1.4 control and eventually the dual end stop option for maximum precision.

My working envelope size will be a minimum 12" x 12" x 3". The reason for this, is that I plan on machining metals. Recently, believe it or not, a few builders have even machined steel with this. This will also allow the mounting on a 24" x 24" table size.

Just a couple of days ago, the designer released another revision with metal parts machining in mind, called the "Burly" set. I'll post a photo of the make when complete, but it may run into next year at the rate I'm going. The designer has sold out of pre-programmed Rambo 1.4 boards, and it may be a while before re-stock.
 
So far, I have the conduit cut and the base parts 3D printed. A couple of parts took 12 hours each. Dimension wise, everything has worked out OK. The bearings all roll smooth as silk, especially if I polish the conduit with 0000 steel wool and then metal polish.

The designer/supplier got in a new batch of Rambo 1.4 boards, so I ordered one immediately. It sits at the ready. I still have to order stepper motors, some 4 conductor wiring, and some special connectors and crimp terminals for the Rambo board.

The photos are just a temporary assembly to check fit and alignment. I will take it apart down to individual pieces again, before mounting it to the operational base.



VdbkjnGC
pySdhqZv
 
Update: I have had the MPCNC operational for a few weeks now, but now I am in the process of "bullet proofing" the wiring with cable chains and sheathing. This has taken a bit of time, as I found I can manufacture my own cable chain for about $3 per meter, as opposed to $9 to $12 per meter bought online. I already have it mounted, aligned, and squared to a fair degree of precision. One of my first projects was to have the machine do a self-portrait plot. Once I get this thing ready for actual milling, then out to the garage with it. For that, I still need to build a suitable table platform.

I used EstlCAM (free to evaluate and try) to convert a .dxf file to a .gcode file in a certain sequence. That is a whole other learning experience. While I am familiar with G-code from my 3D printing exploits, going to the 3D machining is another matter.

One photo has the plot in two different sizes. That was an experiment in scaling the G-code from the .dxf file. A bit like scaling in a 3D slicer, but again with differences.



 
Back
Top