CNC Plasma Build

Over the last 2 years I have been planning to build my own CNC plasma cutter. Various obstacles have hindered my efforts, but somehow I never let go of the idea! About 6 months ago I sat down and weighed up my options and decided to go for it, so since then its been a project that is taking priority.

I now see this type of machine to be a healthy investment as well as a major advantage in terms of creativity, flexibility and productivity.

For those of you who have no idea what a CNC (Computer Numerical Control) plasma cutter is, then the video below will give you some idea of what I'm hoping to achieve.



The machine in basic terms is a 2/3 axis machine that cuts 2D shapes from sheet metals, controlled by a computer that follows a set of coordinates and commands. Usually these coordinates and commands are generated from computer aided designs.  There are a few variations of these types of machine but this is the general idea.

For the purpose of a CNC plasma cutter we will talk about a 3 axis machine:

 Take for example the 3 axis CNC router above. There are 3 degrees of movement X Y and Z, which when instructed to move within set parameters will perform a given operation on a workpiece. On a plasma cutting machine the x and y provide movement for the coordinates of a 2D shape, while unlike a router the z axis does not control the depth of the cut. Typically plasma cutting is a through cutting operation and so the movement in the z axis controls the quality of the cut, which we will cover in more detail later.

Where to start?

 

Exactly!! I had major difficulties deciding what type of machine to base mine on, what mechanics I should incorporate, how big to make it and so on... In the end it all came down to time and money. I opted to design for the best quality and accuracy I could afford. So 2 bog rolls and a paper clip it is then! But seriously, I have a challenge on my hands. 

I began by researching for hours and hours into different professional and diy builds. I visited a few local companies that use this type of machine in Industry and I got swamped with information. Some of the machines cost tens of thousands of pounds!! So after much consideration I selected elements of the existing examples I had seen and put some of these methods as well as some of my own inventions into a hypothetical design.

Generally, I broke down the elements of the machine and looked at each part separately. I then researched further into what existing mechanical products in terms of the 'nuts and bolts' already existed and tried to design around these. My aim was to incorporate off the shelf components to minimise cost and design time.

Drawings, Drawing Drawings...

 

I love my imac, but for some reason the CAD world doesn't. Finding CAD software for an Apple computer wasn't easy. Of course autocad has a version for Mac, but I don't have that kind of cash to spend on software. I looked around and liked the look of TurboCad for mac. The price wasn't too bad, but as luck would have it there was an unused copy for sale on eBay, boom, bargain!!

TurboCad took some getting used to and I'm still learning how to use certain things now. Primarily I've been concentrating on 2D drawings that will be translated into working drawings to produce laser cut parts.

Drawing A

Drawing B

The above drawings were used to produce these custom parts below:

Cut from 8mm Aluminium plate

These parts will form part of the motor pulley assembly for the X axis. The Gantry on my machine will be dual driven by 2 x Nema 23 Stepper motors with a belt reduction of 3:1. The resulting output will drive a 17 tooth spur gear along a rack to provide linear motion.

Sourcing Parts:

 

Some of the components for the drive assemblies on the X and Y axis.

Whilst completing various drawings I have been trying to make sure I keep ordering some of the required parts every month. By doing this its easier to manage my cash flow but keep the project moving.

Sourcing parts is not easy especially whilst keeping to a budget. Many of the required parts are made overseas and therefore some trust in the international suppliers is required.

Left: THK Linear Actuator ( Z axis)     Right: Hiwin Linear bearing blocks

Above are some of the parts that form the linear motion for the machine axis. The THK Linear actuator was sourced from the US and is a complete unit comprising of a precision linear bearing and ball-screw combination. The Hiwin bearing blocks complete with rails (not pictured) form the precision linear travel for the x and y axis. These components are the most expensive items in the build. I felt that smooth travel on the axis was imperative, so I heavily invested in this area.

I will update the progress on these assemblies as the build progresses.

Building the main framework:

I've played around with the size and construction of the main framework and finally came up with a frame that I hope will be rigid enough. It is designed to be built in sections to make moving the machine easier. I have also tried to design an assembly that has features for minor adjustments to allow the precision rails and rack gears to be aligned more easily.

Steel sections cut to length

Pre cut laser profiled plates

Upright legs drilled to allow horizontal beams to be bolted through.

Once I had tacked all of the sub assemblies and beams I loosely bolted the frame together for the first time:

The large tank inside the structure is to store water that will be used beneath the cutting surface. I will cover this in more detail later, but if you observe the video clip at the beginning of this page you will be able to see where the water is used in the cutting operation.

 X axis:

I've spent the last week or so working on the linear motion system that serves as the X axis on this machine. The components that make up this axis are precision Hiwin linear profile rails and precision grade MOD 1.5 racks on each side of the cutting table. These components are mounted in steel channel to help minimise water and dust ingress from the cutting operation and also to prevent accidental damage during the loading and unloading of sheet materials onto the table. Onto the Hiwin rails are mounted 4 bearing blocks (2 for each side). These then support custom laser profiled plates, which form the upright structure of the gantry and are also used to mount the motor and pulley assemblies.

Profiled side plates with bearing blocks for the gantry.

 The rail sits slightly recessed from the edge of the steel channel, so in order to mount the side plates  custom aluminium spacers had to be machined. This can be seen between the bearing blocks and the plate in the picture above.

Assembled onto the linear rail.

Detailed view of the linear rail and rack gear mounted in the steel channel.

Motor and pulley assemblies (X axis):

Kind of in the background, between working on the main framework and other activities I have been making progress on the motor and pulley assemblies. These are constructed from the aluminium profiled components featured further up the page. As previously mentioned, the plates were cut from 8mm aluminium by an industrial laser cutter. However, there was still a certain amount of machining needed to be done before they could be assembled. A local friend and experienced engineer provided the machining service and produced results far beyond my current abilities.

The loosely assembled motor pulley system is pictured below:

View face on

Side view

stepper motor and pinion gear.

The basic principle is this:

The high torque nema 23 motor turns a 20 tooth pulley, which then drives a 60 tooth pulley via a belt. This produces a 3:1 reduction ratio, i.e, 3 revolutions of the smaller pulley results in 1 revolution of the larger pulley. The idea behind this is to trade off the speed of the motor and create more torque and greater resolution. The larger pulley is then connected to a pinion gear via a shaft. This gear will mesh with the rack on the x axis to create linear motion. The rack and pinion system is a step up system so some of the speed lost in the pulley reduction is gained back along with the desired torque and resolution. The pinion has a diametrical pitch of 25.5mm or roughly 1". So 25.5 multiplied by Pi = 80.11mm or roughly 3" circumference. This means that a single revolution from the output of the pulley reduction will move the pinion gear along the rack roughly 3 inches. I have chosen inches because I'd like to measure the speed of my table in Inches Per Minute (IPM).

Control systems:

 

The control system I will be using is based around the mach3 software from ArtSoft. This has proven to be a very popular platform with many CNC builders because of its flexibility, ease of use and strong support network. The software functions as a PC based CNC controller that controls the attributes of a machine through a computer's ports (often parallel port). Lots of small companies have developed products that interface with mach3 as a kind of off the shelf solution for diy machine builders. This means that there is a ocean of hardware out there such as break out boards (BOB's) and motion controllers that allow a user to get their system up and running quickly. I've opted for a motion control system by CandCNC, FourhillsDesign USA.  The product they supplied is the MP3000E - DTHCII, which is an all in one motion control system for a plasma cutting table. This incorporates a digital torch height controller (DTHC). I will cover torch height control later. The MP3000E - DTHCII comes in the form of a compact enclosure, which has separate control modules linked remotely through generic control cables. These modules provide access to step and direction pins, optically isolated digital input and outputs and a plasma machine interface, which are all accessible through Mach3 using the custom screen sets supplied by CandCNC.

Previously I had toyed with the idea of designing and building my own control station to operate my machine from. I drew up a concept to see what my ides would look like:

Front panel

Panel mounted in an enclosure

While I was pleased with my concepts, I needed an instant solution that required little design and build time because of my commitment to the actual machine build. I looked at various products that already existed, but one that seemed to provide an effective, pre built solution was the touch screen kiosk. These are the types of machine often found in museums or large shopping complexes that help visitors access information about the building or exhibits. The touch screen systems are built to high specifications, often incorporating peripherals that are hard wearing and long lasting, perfect for the harsh environment of a metalworking shop.  Some brief eBaying and the perfect candidate was bought, packaged and delivered. Job done! almost.....

Touch screen system running mach3

Side view, showing the sleek design.

View from the back with panels removed showing industrial pc and motion controller installed.

With a control solution sorted and work on the axis nearly complete it was time to have a play with all of the items mentioned above and to see if I could get the thing to move. After a day of wiring, testing then re-wiring I was able to move the X axis with mach3 in jog mode. Success!! At the end of the day my test bench looked like this:

excessive use of computers and cabling, garnished with a cup of tea :)

  Also during my week spent on control systems and electronics I built up the basics for the control enclosure that sits in the main frame of the plasma table:

Control enclosure, still requires some items and needs wiring.

This box will contain all of the motor controllers, power supplies, input and output modules with all necessary switchgear and protection devices. Once mounted to the main table it will serve as the 'guts' of the machine.

 X Axis progress:

After a few weeks of overcoming slight problems with the X axis drive assembly I now have it working quite well. The problems I was faced with were caused by some of the mechanical elements surrounding the tension system for the rack gear. Due to slight foreseen discrepancies in the rack gear as the pinion travels along means that a mechanical system that keeps the pinion under constant tension needed to be designed. My Initial design included a number of springs to supply the pressure to overcome this issue. Despite testing many different types of springs in this system the problem still remained. I then started to look at other simpler solutions and sourced some gas struts. These provided a compact and tidy solution, also I was able to specify the required force upon ordering. Once assembled the gas struts applied enough pressure from the pinion to the rack along the entire length of the machine to help eliminate backlash. I was then faced with a separate issue, which was the sliding mechanism that allows the pinion gear enough give to keep even contact with the rack. I had allowed too much slack, which meant I had to get these re-machined and have bushes made to the correct size.

X Axis motor assembly.

The picture above shows the motor / pulley assembly in position. The gas struts can be seen either side of the belt. These work together to provide pressure to allow even contact between the drive pinion and rack gear along the length of the machine, helping to eliminate backlash. The gas struts force the motor / pulley assembly upwards, which is guided by slotted bolt holes.

Y Axis:

 

Work on the Y Axis is well under way and will be ready for troubleshooting in the next few days...

Y Axis motor / pulley assembly

Y Axis and gantry

Gantry

New compressor:

With my plasma build coming on I need to start planning for the accompanying equipment, which will support it. A fundamental element of the plasma cutting operation is a good quality air supply. I already had a large piston type compressor, but this is noisy and cuts in and out all of the time to try and compete with the supply of air the plasma machine needs. Therefore opted for a more robust system, a screw compressor. Screw compressors work differently to standard piston types and are able to supply compressed air in much higher volumes. They are however, more complicated and much more expensive to buy, run and service. I bought a used screw compressor with very few hours on it. I had it fully serviced and commissioned and is now ready to run :)

Screw compressor, dryer and receiver.

 
The plasma cutting process requires a clean dry air supply. Therefore, an air dryer and relevant filters are required. The blue machine in the picture is the dryer in my system, which I have yet to pipe up.

Motion testing:

 

Once the Y and Z axis were fitted to the gantry I had to set up the software to send the correct signals to the drives in order to move all axis within the same units the software was measuring. This got quite complicated as there are quite a few mechanical drive elements involved. Basically, I needed to make sure that when the software told the motors to move a set distance i,e 20mm, the machine moved precisely 20mm. It took quite a few hours of calculations and cups of tea to get this correct. Currently the machine holds an accuracy tolerance of 0.025mm, which I'm hoping to improve by correctly tightening mechanical driving components.

I found a test file online that would draw the shape of a scorpion and decided to have a play with the machine:

Drawing with a pen onto cardboard.

The machine is far from finished and I'm yet to get my head around the software completely, but I was so excited to see something I had build from scratch, running independently.

Y and Z assembly.

Above shows a closeup of the y and z assembly complete with the torch holder. You can see from the picture quite a few micro-switches. These switches are limit switches that tell the control system to stop once the machine has reached the end of its travel and to prevent damage to itself. There are more switches still needed, these will be home switches. The home switches tell the machine where its 'home' position is, often positioned within the machines safe working envelope.

Updates:

So its been a while....

This build has been sitting quietly in the background for a bit, while I've been concentrating on various other projects as well as trying to do work to fund everything.

The machine is nearing its finished state now and because there is a backlog of build photos, I thought I'd add a bulk load of images up to the state I'm at now:

Thank you for taking the time to read this page. The machine is now complete and making parts everyday. For a little side project involving parts made using this machine please please follow my twitter page : https://twitter.com/EngineHouseGB