Embedded Files
Build One
It sounds like the model that the OSE guys are working on will be really cool but it is not ready to go yet as they have not drafted the design or built a prototype yet.
The following video is from November 2019 and discusses where they are with the torch table at this point. I have summarized the key points from the video below:
- their end goal is to produce a workable kit
- at this point the OSE people are still outsourcing their CNC cutting
- to purchase a model they estimate would cost about $10k and they are trying to design something that you could build for about $1k
- they like using oxy-acetylene because you can get deep cuts for a relatively cheap setup
- in addition to the 3D printer the Marlin/Lulzbot Cura toolchain, they are also using this software for the CNC torch table as well as Flatcam
- you can start by drawing in FreeCAD and export the design as a .DXF file or you can start in Inkscape however FreeCAD would probably work better for precise shapes
- draw in FreeCAD and export .DXF file --> import .DXF file in Flatcam and export .GCODE --> import .GCODE into the Universal Controller using an SD card or a USB port
- CNC torch table geneology
- Prototype 1
- They used it to cut parts for the tractor
- This version used Linux CNC using the serial port but few computers have that type of port
- Prototype 2
- aborted version that they never built
- Prototype 3 (version 17.08)
- they called this prototype the D3D CNC torch table
- aborted version that they never used but they did build it
- this version was a 5' x 10' structure with 1" rods and tiny NEMA motors
- they never used it because the structure was not stable enough
- the 1" universal axis can go about 5' but beyond this length the axis begins to sag and wobble
- they proposed using 1" schedule 40 pipe to reduce the weight of the axis and therefore the sagging
- the autoheight correction may correct for any sag in the axis but wouldn't fix the stability of this prototype
- Prototype 4
- aborted version that they never used
- Version 19.10
- they are still working on the design of this and drawing it up in CAD and it still needs to be built/tested
- uses a 1" universal axis
- they have a google slides presentation entitled "CNC Torch Table v19.10"
- designed as a 5' x 5' table with a water bed that is modular so you could connect two of these together with minimal effort to make a 5' x 10' version
- axis was sag in the 5' x 10' version but they use the same proximity sensor that they use on the 3D printer to make a height correction but you can't go fast otherwise it will wobble so they suggest putting in a support angle under the carriage (assembly that holds the torch) and a roller on the bottom of the carriage so the carriage glides on the angle support
- the roller would be protected from splatter since it would be outside of the frame
- axis was sag in the 5' x 10' version but they use the same proximity sensor that they use on the 3D printer to make a height correction but you can't go fast otherwise it will wobble so they suggest putting in a support angle under the carriage (assembly that holds the torch) and a roller on the bottom of the carriage so the carriage glides on the angle support
- he guessed that the 5' x 5' table would weigh between about 500 and 1000 lbs
- he is designing the axis so it hangs from the frame to avoid 1" holes since these holes are difficult to produce and to get at exactly the right place and allow for the table to be modular
- they are moving in this direction with the 3D printer so they an be moved along the vertical axis
- this version would also control the flow of gas with a ramps 1.4 board and a acetylene solenoid, oxygen solenoid, and a cutting solenoid so it can work with various options including only acetylene with ignitor, heating mode, and cutting mode. this feature will also allow you to turn off the gas as the torch is moving but you don't want it to be cutting
- to do this you will need a 3 tube positive pressure torch which has a tube for the acetylene, a tube for the cutting oxygen, and a tube for the mixing oxygen that is modified for this functionality
- Prototype 1
- he also suggested a goal or changing from an oxy-acetylene setup to an oxy-hydrogen setup so you could make your fuel by splitting water via electrolysis
I also found instructions for building one here: https://www.instructables.com/id/CNC-Plasma-Table/
This guy created a series of youtube videos showing how he made his CNC plasma cutting table: https://www.youtube.com/playlist?list=PL4lExYUFHSPy926ACka0Rf6WGTWqWWXS4
This guy shows a bunch of pictures of him building it but doesn't really give the details needed for replication but it could be used for some ideas: https://imgur.com/gallery/JitIU
Buy One
The model from https://www.langmuirsystems.com/ with a 2'x2' working surface is about $1600 for the table and then about $800 for the plasma cutter. It doesn't look like we could alternatively use our oxy-acetylene torch.
This model from https://millrightcnc.com/product/mega-v-plasma/ is about $2,350 has about a 3'x3' working surface.
Kits and/or tables can be purchased from http://www.cncrouterparts.com/. Just the electronics kit runs around $1,100.
Using their kit and building your own table, a 5'x10' plasma table would run about $7,500 as shown here: http://www.precisionplasmallc.com/default.htm.
In a forum someone from hypertherm recommended using the electronics from here: https://www.candcnc.net/ and the mechanical parts from here: http://www.precisionplasmallc.com/ .
Rather than buy an entire table, we can build the table and buy the CNC part. I've listed a comparison below. The cheapest options appear to be through Alibaba but if we don't feel comportable with that then we can try going through Amazon.
Outsource It
We could alternatively outsource this work and see what it would take to get the guys at Pacific Steel and Recycling to cut it with their equipment or a vendor online.
Decision
At this point we are planning to wait for OSE to prototype the version 19.10 model and release documentation and if it appears that their model is working well then we can build it. There is some existing documentation so we could potentially get a jump on the documentation that is available however this comes with the risk that this model doesn't work out well or they change the design. Marcin said that he expects documentation to be available after this summer.
Google Sites
Report abuse