Monday, 29 July 2013


Theory & the basics of the design

Mechanical Design

The “chassis” of the machine has been built from a single full-size sheet of 18mm MDF and divided in such a way as to reduce the cutting and to use the already, reasonably square edges left by the board makers.  The linear bearings used on the axis’s have been sourced from kitchen draws, granted the accuracy of these runners isn’t the best, but by installing them in such a way that they are either under compression or being opposed, deflection can be minimised.  These draw slides have been chosen as they are cheap and easy and not because they are accurate or suited to the task in hand!  The threaded rod is M8 x 1.25mm all-thread bar and again this is chosen as it is cheap and easy to acquire, not because it is accurate or ideally suited to my application.

Spindle

To keep costs down, for the spindle (cutting motor) I’ve used a cheap dremel syle rotary tool.  This has a 3.2mm collet fitted as standard to allow the use of small tooling and spins up to 35’000RPM with 130 Watts of power.  Obviously this isn’t as powerful as the motors found in commercial machines but it’s cheap and allows scope for improvement at a later date.  The multi-tools high speed is ideally suited to the small cutting tools used, which must be spun at a much higher speed than larger diameter tools.  The tool also has an in-built speed control, allowing the spindle speed to be adjusted manually.

An important point to consider when choosing one of these tools to use as a cutting spindle is the amount of run-out at the tool tip, this can be seen as the “wobble” of the tool as it turns.  This run-out will cause inaccuracies in the machining of the part but will also induce a radius of gyration and cause a vibration in the machine.  This vibration will show as “tool-chatter” and leave marks on the parts surface.

Once the machine is up and running I’m going to be looking at making a new spindle from a Radio Control Model Plane DC Motor and a ER11 spindle extension, but that can wait!

Tooling

Due to the 3.2mm collet, there is an awful lot of cutting bits and burrs available from the internet for next to nothing! I bought a 3mm carbide end mill from eBay for £2 delivered and this has been used for most of the cutting so far, as I’m only cutting MDF and soft materials, this end mill should last a pretty long time.  I’ve also got some “V-Carve” tools to try and some 3mm drill bits as well as some larger 6mm ball end mills for when I get chance to develop the new spindle.

Control System

The control system to be used is the tried and tested stepper motors via serial com port and a driver board.  The system works in the following way; firstly a model is generated in CAD of the finished part.  This is then exported into a CAM program which generates “G-Code”, this G-Code data is transferred to the machine controller software which translates this G-Code into step and direction pulses out of the serial coms port of the PC (at 5V and very low current).  These signals are sent down a serial cable to the stepper motor driver board which uses them to drive the stepper motors by energizing the stepper coils in such a manner that they move the required amount.  The power supply to the stepper motor driver board comes from an old PC ATX power supply and these supplies the required current to the board to enable it to drive the stepper motors.

As you can see above, the opportunity for errors in the finished part can creep in to the system in many areas, the main one being the computer running the control software is required to generate “pulses” for the step and direction signal, if these pulses vary in there speed then the machine will vary the distance it travels for each one, that is to say that if the frequency of these pulses alters from the expected frequency then the machine could travel at a distance of less or more than expected and the finished part’s size will vary away from that expected.  The amount of this frequency deviation will vary on the “load” on the computers CPU whilst running the “heartbeat” algorithm that controls this output frequency.  For instance starting a program whilst the PC controller is running could cause a load on the CPU and an error in the machined parts finished size!

There could also be an error induced if the parameters of the machine are not correctly entered into the machine controller software, if for instance the leadscrew pitch was entered as 1.2mm instead of 1.25mm the machine would assume that to travel 50mm it would need to rotate the axis 41.666 times as opposed to 40 and this would give and error of 2.08mm on a 50mm cut!

Monday, 22 July 2013


Some background on me and on the project (oh and my first blog post!)

This blog is intended as a way of sharing my interest in home-made CNC machines and machining with the wider world.  I’ve been interested in CNC milling for years after seeing some pretty interesting videos on YouTube a few years ago, having done some manual milling whilst doing my apprenticeship I could understand just how amazing these machines actually were, and how difficult the tasks they were performing would be to accomplish.  I kept an eye on the technology whilst doing the first part of my degree and this turned into doing a placement year with a DELCAM as part of my degree course, for those that don’t know, DELCAM are like the Rolls-Royce of CADCAM software and are the biggest supplier and experts in this sort of software.

During my year with DELCAM I got to spend some time in their on-site machine shop looking at everything from micro-machining of teeth to machining huge pieces of stone with a robot cell (as well as the 5-axis aerospace parts!).  When I went back to University to do my last year of my degree course, I designed and built a small desktop CNC from aluminium and stepper motors, luckily I didn’t have to pay for the materials as it ended up costing around £500.  This project paved the way for the work seen in this blog and gave me some really great ideas to reduce costs and simplify the build process.  Unfortunately when I graduated I joined another company and lost access to any sort of CNC equipment, so project “Super-cheap-CNC-mill” came into action!

The project covers the construction and build of what I believe to be the cheapest and more importantly easiest CNC milling machine to have been documented on the internet and everything used has been chosen to be CHEAP and EASY to use, therefore there are LOTS of areas where there could be improvements in performance, so please don’t bear this in mind before giving one of the usual “you should have used ball screws and linear bearing” type comments!  I gathered all of the material together and made a start on Saturday morning, by Saturday evening I was running my first piece of code! That’s 8 hours to build!