Archive for the ‘CNC Plasma’ Category


I am never short on things to fill my time with and typically I need to implement time management strategies in order to accomplish the things that I consider to be important. One of the things that suffer is the time spent on the internet looking at other cool projects people are doing. There is soooo much stuff out there that people are doing that it actually frustrates me because it inspires new ideas, and projects, that I do not have the time to take on. I am always intrigued by the blogs, and sites, that show close up photos of the actual work that is being performed and not just the finished projects. I feel as though, over time, my blog has lacked the visuals that provide the raw metal and tools. This posting I wish to get back into, what I consider, to be the passion.

For those who need an update on what is going on in the garage these days I am building a plasma torch CNC table. Check out my previous posts to get up to speed if required. This blog entry is going to deal with the Z axis. As stated previously I am building the table “backwards” and starting from the tip of the torch.

The Z axis dose have some requirements. In my case I am designing a floating torch head. For those not familiar with this style I will briefly explain. In order for the CNC software to know the vertical position of the torch head the Z axis stepper motor needs to run the torch head down until it touches metal. There are numerous ways for the software to know when the touch occurs. In my case the torch is designed to touch the work piece and then it will start to float on the Z axis. In other words the torch head stops moving once it touches metal however the Z axis continues to travel downwards until a mechanical switch is triggered. Once this switch is triggered the software can then back the Z axis upwards to a pre-programmed dimension which will set the torch head at the proper starting height. The reason for the floating design is to prevent any damaged that may occur to the torch head while being set onto the work piece. Stepper motors have enough force to start breaking components of the table and torch. By only floating the head the weight the torch head, and support plate, is being exposed to the torch itself.

I did AutoCAD some basic starting points in order to machine my main support plate. After the initial fabrication of the plate took place I started to just wing it all. The following pictures show the process I used to create a mocked up version of my Z axis. Pleased to say I hooked it up to the power supply and PC and was able to run it through its vertical motions will no issues. So at this point the Z axis is tested and working. There is still much “clean up” to do on the parts including trimming of excess aluminum. I will do this at finishing stage.

The way I post most of my pictures, for other blog entries, is in sequential order. I start from the beginning of the project and finish at the end. I changed things slightly this time. Since there are multiple smaller parts that make up the entire Z axis I tried to start certain sections with the finished part. I am hoping that it may help those, who are interested, in following along with the pictures a little better. There are 43 pictures posted in this one, most of them machining shots. If you don’t understand what’s going on I encourage you to take comfort in the visuals of 6061 aluminum, spinning tools, and flying chips.


This is the roughed out Z axis with a floating torch head. None of the finishing details have been addressed plus much of the hardware that holds it all together is not installed. The following pictures outline some of the processes I used to build it.



My arms got a work out using my manually operated draw bar to swap my tooling in and out of the milling machine. Part of the purpose of building the CNC is to give me some more “mill time” to get better at using the tooling.



I AutoCAD’d the basics and then eventually got to a point where I just started winging it. This is the backing plate getting drilled and tapped according to my CAD specs. The plate is the backbone of the entire drive which will be used to support the X bearings, the X drive, and the entire Z axis floating torch head assembly. It’s built from 6″ x .375″ 6061 aluminuim.



The ball screw that drives the Z axis nees to be support by 2 sealed ball bearings. Using the boring head attachment I machined a press fit hole in some 6061.



Holes were drilled and tapped in order to bolt the bearing flange onto the backing plate.



Next came the pressing in of the ball bearing assemblies. A vise and a socket works just fine in order to squeeze the 2 parts together.



I required a coupler in order to connect my stepper motor to the ball screw. This is the final, rough machined, product. Since my ball screw shaft size differed slightly from my stepper motor shaft size I made the coupler with 2 different bore sizes on each end. The following few pictures shows the machinig of the coupler.



After spinning down some 6061 solid road of aluminum on the lathe and drilling my 2 different sized holes I moved onto the mill. First step was to mill pockets to accept the heads of the socket head pinch bolts.



Next the coupler was drilled and tapped for 5mm stainless steel hardware.



Final step was to make it into an actual coupler by running the slitting saw through one side.



Here you can get an overview of how things are starting to fit together, The bearing supports are bolted to the backing plate and the stepper motor is installed and coupled to the ball screw.



Moving along I needed to build the floating head portion as well a clamp to secure the plasma torch head . A strain relief for the plasma cable also needs to be integrated. Here is a shot of what the following pictures created.



The Hypertherm Powermax 45 torch head has a 1 inch diameter to clamp to. I started with a block of 2″ x 2″ x 1.5″(?) 6061 and just started milling. First step was to bore out the hole to give a nice, clean, slide fit for the torch head.



The plan was to make it a pinch style clamp. Holes were drilled and tapped for 2 pinch bolts. I performed step drilling and then only tapped the bottom half of the holes.



In order to allow the clamp to bolt to the floating torch head backing plate more holes were drilled and tapped.



I needed to build some flex into the clamping blocking in order to ensure the torch head will clamp securely. I milled off some of the side material in order to thin the block up and allow flex.



Final step was to run a slitting saw through it. The slit seperated the drilled bolt holes from the tapped holes.



Next step was to fabricate some sort of strain relief to support the torch cable. No plan here, just picked up a chunk of 6061 and started to remove metal.



Started on the lathe by machining the center hole to the dimensions of the torch cable OD



Figured I would pretty up the holder so before I removed the part from the lathe a ran a decorative groove into it with the part off blade.



Onto the milling machine. The strain relief will be secured with four 5mm stainless steel socket head screws. I milled some pockets into the clamp to accept the screw heads.



Next all four holes were drilled and tapped.



Time to split the strain relief in half so that the torch cable can be secured in it.



A shot of the rough machining completed so far.



The strain relief needs to be mounted to a top plate and therefore one side of the strain relief needs to be shorter then the other. Milled off .375″ on the rear half.



More holes were drilled, and tapped, in order to allow for mounting to the suppot plate.



Top plate was drilled with the same radius in order to accpet the strain relief.



I have a .250″ ball nose endmill that I never use for anything. I thought I would add some useless detail to the strain relief.



This is the rough machined strain relief mounted to the top support plate. I may anodize it orange, along with the torch head clamp.



The following set of pictures are mostly random shots of the machining used for the remainder of the z axis. This shot is the start of the coupling block that will get bolted directly to the ball screw. This coupler will then serve as a means to connect the linear bearings to the ball screw. This is a 2″ x 2″ chunk of 6061 getting milled to make everything square.



Next I roughed out an opening with a 1″ drill bit.



Holes were drilled and tapped to match the bolt pattern of the ball screw nut.



The roughed 1″ hole got final machined to the dimensions of the ball screw nut.



Finally the coupler was milled to dimemsion in order to allow it to operate on the same plane as the linear bearings.



Random shot of the coupling plate. This plate gets bolted to the coupler shown in the previous 3 pictures and then is connected to the linear bearings.



I machined some UHMW (Ultra-high-molecular-weight polyethylene) to act as a slider to support the floating head plate.



Once I am sure everyting is operating to spec I plan to then start trimming off excess aluminum. Here the stepper motor support plate gets a rough trim to shave of some weight.



Next it gets cleaned up on the belt sander. The corner radiuses do not require precision.



The next 3 shots are showing the assembled components in rough form. The stepper motors I am using are dual shaft. The motor has a different diameter shaft on each end. Since I am only using 1 shaft I decided to machine up an aluminum disc to spin on the top shaft. I will probably anodize it and sandblast a logo into it yet.



Here is the side view of the floating torch head. I realize it is hard to understand the mechanics of it without more shots, video, or getting your hands on it.



This is the mechanical switch I mounted in order for the software to detect the starting position of the floating torch head.



It was time to clean out the mill, lathe, and the floor.



Here are all the components that make up the z-axis. So far this is all just rough machined. Sill have lots of clean up to do plus the finishing.



The plasma table CNC build has officially gone into fabrication phase. I have sacrificed multiple, sleepless, nights coming up with a game plan and determining the best sequence to build the table in. I have opted not to use any existing plans but instead engineer the table my way. I have a “big picture” in mind however all the details that are required to ensure the concept will be completed are yet to be determined.

Most tables are typically started by building the main frame. Since there are so many unknowns as far as gantry sizes and, more importantly, X,Y,Z travel dimensions I decided I am going to start from the center of the universe. In this case I am going to build the entire table around the tip of the plasma torch. This would mean I begin with my Z axis.

Since part of the construction of the Z-axis involves its ability to move along the X-axis I needed to come up with a linear movement system. X and Y axis linear movement methods are obviously nothing new. There are multiple systems that a proven to work well. My favorite has always been the Dualvee Bearing design coupled with the Vee rails. It takes care of both the radial and axial movements all in one shot and it does it in a fairly compact set up.

Since the whole point of fabricating the CNC is to actual “fabricate” I wanted to avoid purchasing as many components as possible and instead build the items. Coming up with a simple linear motion system that I had the skills, and equipment, to build was tough. I didn’t want to clutter up the sliders with aluminum plates housing 8 bearings each just to keep things smooth and straight.

After much thought I took some inspiration from the Vee bearing design and opted to build my own version but without the Vee. My version would incorporate a radius bearing that would ride along a 4140 alloy rod. If the design works it will control the radial and axial loads just like a Vee bearing does.

Weight of the table is a huge factor and this will become evident why later on in the build. So after a bit of experimenting I came up with a system to accurately machine radius bearings out of aluminum. Aluminum is not exactly the first choice for bearing material however in my case the loads are not massive plus the ability to anodize aluminum will certainly add a layer of hardness. The following outlines the first steps in building the CNC table by starting with the axis bearings.


People often ask for the plans of some of the things I build. I figured I would start by posting the intricate CAD drawings of the bearing assemblies. Here is all the information for the world to see.


Here are the bearings in various stages of production. They are all machined from 2″ 6061 round bar. The outer finished dimensions are .750″ wide with a 1.975″ diameter.


The blank is hogged out with a 1″ drill bit on the lathe.


Next it gets bored out down to the .001″ to accept the press fit bearing.


A custom fabricated arbor was required in order to perform the external machining using both the lathe and the milling machine. Here is the steel arbor I built to secure the aluminum blanks.


The outside diameter gets lathed down to 1.975″ before it gets moved over to the milling machine.


The rotary table was set up vertically and the blank gets secured in order to allow a .125″ deep cut using a .675″ endmill.


This is the final machined bearing housing before it goes into finishing stage.


All the bearing housing then got a 2 stage polishing in order to smooth things up.


Here is the set of 8 ready to move on to the anodizing phase.


In order to remove all of the cutting oil and polishing compounds the housings recieve a soaking in a heated solution of SP degreaser.


After getting a good scrubbing in hot soapy water the units get rigged for hanging in the sulphuric acid anodizing bath.


Everything gets hung and electrically connected ready for a 2 hour soaking.


The power supply gets connected to the bearing housings and 4.50 amps at 15 volts is dialed in for 120 minutes.


These are what the housings look like after they are anodized. The shiney, polished, aluminum turns to a dull light grey color.


All the housings then recieve a dip in the heated orange dye tank. To keep the color consistant they are all timed for an 8 minute bath.


Here are the housings fresh out of the dye tank.


Time to press the bearings in. In order not to damage the finish on the housings I machined a couple of bushings that I used to keep the housing, and bearing, straight and protected while they get mated using the vise.


Finished product. I will be curious to know how well the anodized surfaces will stand up to wear and tear. Having them colored orange will make it easy to determine the extent of wear they are suffering from.


So winter is approaching which always signals time for change. As the colder temperature and the snow begin to sink in as an unstoppable reality the planning for hibernation is inevitable. The past couple weekends have been spent putting the yard to bed. Winter fertilizer has been applied, trees and bushes have been pruned, sprinkler system has been blown out, and the gas yard equipment is set for winter servicing and storage. Now what!!!!??????


Last winter I was able to occupy my time completing my 1965 CB160 Café Racer build. It was a good project that resulted in great success. It allowed me to enjoy 2000km worth of riding this summer which included participation in the 2014 Distinguished Gentleman’s Ride.

For the winter season of 2014-2015 a new project is in order. I have many ideas stored in my database, the one that rests upon my neck, and it was just a matter of choosing something. Over the past 4 years I have explored the idea of building a CNC plasma table. I did much research many years ago which mostly involved reading manuals surrounding the CAD, CAM, and CNC operation and design. After having a few years to let the massive amount of information sink in I decided this winter season is the time to make this project happen.

The details involved in all the pre-planning are substantial. I am not going to bore you with the entire game plan but instead I will let it all unfold in the blog. So for today I will simply set out the ground work which is necessary in order to understand what direction things are headed (yes…pun).

First off, if you do not know what a CNC plasma table is then you can read about it here, or you can just Google it. Not going to cover the basics in this blog. What I will cover is the basic equipment involved in building the machine.


First piece of equipment involves an actual plasma cutter. In my case I will be using my Hypertherm Powermax 45 unit which I purchased many years ago. I settled onto the Powermax45 for a number of different reasons and 1 significant reason was the ability to interface it with CNC to allow for torch height control. This unit is spec’d to cut up to 1” thick steel and can do .500” at 20 ipm (inches per minute) I can tell you, from experience, this unit is a work horse. Creating plasma and molten metal is what it does best!


Next important piece of equipment in the CNC table build involves the tables X,Y,Z axis movement control. Basically this involves some stepper motors attached to the tables gantry’s that allow for computer controlled movement of the torch. After much research I settled on CandCNC BladeRunner Dragon-Cut set up. There are way too many details to list about this unit, you can visit the link if you want to read up on it all. The highlights involve a power supply to run four 620 oz. stepper motors to move the X, Y, and Z axis (1 will be slaved to the Y axis). The Bladerunner also allows for Digital Torch Height Control (DTHC) which assists in the precision of the cut.


Since the BladeRunner is only a power supply controlling some stepper motors it requires a commander. In my case the boss will be a dedicated computer running specific software to shout the orders to the stepper motors. I opted to build my own PC to the specs dictated by CandCNC in order to ensure the Bladerunner will approve of the commanding officer. I stopped in at my local computer supply store and picked myself up a PC case and all the guts needed to get the electrons flowing. The PC spec’d out as follows; Intel Core i3 3.60GHz processor, Asus H81M-E motherboard, Kingston HyperX Fury Black 4GB RAM, Asus 24x DVDRW, coupled with a Kingston V300 120GB solid state hard drive.


I have never actually built a computer and really have no idea what I am doing. I figured as long as I could get all the bits to fit in the box and if I could find a place to plug in all the wires how could it possibly not work? Looks like my theory was correct, I was able to keep all the smoke contained within the components, nothing leaked out, electrons flowed and pixels were produced. The computer will only be running 1 piece of software and nothing else using the Windows 7 32 bit operating system.

So with the power supply, stepper motors, and PC all dealt with the next step was software. The CNC system requires 3 different programs in order to make the magic happen. First off is a design program in order to create whatever it is that will be cut. The second is CAM software which takes the design and turns it into “G-Code”. This code is used to tell the table software where to move the stepper motors. The third piece of the puzzle involves the CNC software. This is the program that takes the G-code from the CAM software and then sends the signals to the power supply and commands the stepper motors. I was not prepared to stray too far from my comfort level so I decided to stick with a guaranteed combination that is known to work. CandCNC recommends certain software which also happens to be some of the most popular stuff used in many homebuilt CNC units. The 4 programs I am using are InkScape design software and DraftSight CAD software as the 2 design programs, SheetCam as CAM software, and Mach3 as my CNC software.


Inkscape is a free version of design software that mimics Adobe Illustrator. I will not go into detail however I will say that it gives my grey matter a major work out. It takes a lot of time to get the hang of the software and I continue to force the education upon myself. There are lots of “clipart” type designs available for CNC purposes which eliminate the need to learn Inkscape however in my case I want to be able to design very specific components. In the case of today’s blog post I am demonstrating a “Gords Garage” gear project going from initial design all the way to cutting. In the above photo I built a gear with a GG in it and vectorized it.


DraftSight is a free version of AutoCAD and works incredibly well. In the case of my GG Gear demonstration I did not use DraftSight. In the picture above I just quickly built a gear to show what is possible. I do not use DraftSight as “artsy” design software but instead use it more for component design.


The next piece of software is SheetCAM. This program takes my vector design and converts it to G-Code. In SheetCAM I can decided how I want the CNC table to cut out my design. I tell it where I want the torch to go, I program lead in and lead out cuts, I decide how fast things should happen, and in what order. Once I have made all the crucial cutting decisions I convert it all to G-code ready to be accepted by the CNC software. I should note that InkScape, DraftSight, and SheetCAM are all run on a separate computer from the CNC therefore after I am done with SheetCAM my G-code is loaded onto a memory stick and transferred to the CNC PC.


Mach3 is the CNC software loaded on the stand alone PC attached to the CNC table. The G-code from the memory stick is loaded into Mach3. I was not about to build the table and then figure out if things work so I have set everything up on the floor in a spare room to ensure I can run mock cuts. The PC, BladeRunner components, including the stepper motors, are all hooked up for testing. With the GG Gear G-Code loaded into Mach3 I can successfully run to cut pattern. In my case all I have is 4 stepper motors buzzing, and spinning, while lying on the floor however this verifies that all my software and hardware is currently functioning. Mach3 also controls my plasma torch height control which I will not be able to configure, or test, until the table is built.

I should perhaps mention that the whole reason I have chosen to build this table is because I like building things in the garage. Everything that I have listed so far does not qualify for enjoyable time spent in the garage. They are, however, necessary components, and steps, required to ensure a successfully completed, and operational, project. A well planned project is a project half done. So now that I have spent 2 months planning, learning, and collecting components I am getting closer to spending some quality garage time.


The actual table build is what this project is about. I have much of the table mentally designed. The plan is to use the tools and equipment I have available to me, in my garage, to build a unique, quality, fully functioning 4’x4’ plasma table. I want to build as much as I can by hand including the gantries, the support system and the floating head Z-axis. I am not basing my build off anyone else’s table but instead going to approach the design using my own logic and ideas. There are many companies that can sell you all the components you need in order to complete a build. In my case I have only purchased the bare essentials and plan to fabricate the remaining. The components I have collected include a ball screw and 2 linear guides for the Z-axis, a rack and pinion set up for the X and Y axis, and a 3:1 timing belt set up for the X and Y axis.

So basically this concludes my introduction to my next project. I have only included a brief overview of what I have started. The blog posts to follow will, hopefully, give you an idea of my direction, design, and completion of the build. At the very least I guarantee to show chips flying and sparks shooting.