Planning Your First DIY CNC Build

What you should know before you start planning to build a CNC…

I wrote this article because looking back I wish I’d read this article before setting out on my dream of building a CNC machine. That’s not to say that I regret building a machine or that I wouldn’t do it again in a heartbeat just that I’d have liked to have gone into it with my eyes a little more open than I did. There are, I feel, three important factors that you have to be realistic about: the materials you want to cut, the size of the largest piece you want to work on and the amount of money you want to spend on the machine. For the sake of keeping this article simple I’m going to write it from the perspective of someone wanting to build what most people think of when they think CNC; a three axis router style machine. Nothing in the name CNC limits a machine to this design but it’s a very common first build.

What materials do you want to cut?

Taking a very broad brush to the materials question I’m going to say there are only four materials you’ll want to cut: foam, wood, aluminium and steel. “Whoa”, I hear you cry, “I want to cut carbon fibre / plastic / stone / unobtanium / etc / etc”. All the materials I can think of require the same basic design of machine as one of the four materials listed initially. Yes, you might need to make certain design modifications (e.g. water proof a stone cutting machine) to cut a more exotic material but the basic principal is going to be the same as one of those first four.

The answer to the question about what you want to cut with the machine is actually dictating what materials you build the machine out of. As a rule of thumb you have to build the machine out of the toughest material you want to cut (except foam, you can’t build a CNC machine out of foam). So, you can build a machine that cuts wood out of wood. Having said that I want to point out that the vast majority of successful CNC projects build the machine mostly out of aluminium. The bed / frame may be steel but all of the important parts tend to be aluminium. Why is this? Simply because building a CNC machine is difficult and using aluminium is the easiest and cheapest way to get a moderately accurate and strong machine.

But I’ve seen this machine built out of MDF…

It’s certainly possible to build a machine out of wood or more likely MDF but the machine will be limited in what it can cut and the accuracy it can achieve. If you only want to cut foam and wood an MDF machine might be for you. If the wood is very hard the machine may not have fantastic accuracy but most of the time it should be good enough. The appeal of MDF machines to most DIY builders is the cost of MDF and how easy it is to work with commonly available tools. To some extent though this is wrong headed thinking, MDF is certainly cheaper than aluminium but if you are going to use real guide rails rather than home built versions, decent control systems, etc, etc, then the cost difference starts to look smaller. A second appeal of MDF to the new builder is that it’s lighter than aluminium, again that’s true but you invariably end up using more of it to achieve a reasonable stiffness so the weight savings aren’t fantastic. Additionally weight is, up to a point, your friend. A heavy machine will vibrate less while cutting which will result in a better finish on the piece.

The question every new builder asks is “can I cut aluminium with an MDF machine?” and the answer is yes with a massive caveat. The milling cutter will go through the aluminium you are trying to cut but the end result won’t be very good. The comparatively low stiffness of the MDF will result in deflection as the tool contacts the aluminium which will result in inaccuracy in the cut and chatter marks on the piece.

So I should build my machine out of steel?

If you can completely build your machine from steel that’s great but if you have the facilities to accurately work steel you probably don’t need to build your own CNC machine! Aluminium can, just about, be worked using woodworking tools. Steel, in contrast, is tough and needs dedicated metal working tools. A lot of DIY CNC builders start out with workshops equipped for woodworking so steel working is next to impossible. One big problem people face is surfacing the steel, the vast majority of steel you buy won’t be flat enough to use in critical areas so it will need milling to produce one or two flat faces to work from. Unless you know someone with a large milling machine they are willing to let you use it’s going to get expensive quickly to get all the steel milled. CNC machines invariably require complex parts as well which will also require a milling machine (or laser cutter / water jet cutter / etc) so again the expense quickly adds up.

The upside of steel is that it’s easy to join by welding. With a cheap MIG or stick welder anyone can join two pieces of steel together with a strong joint. The great thing is that it doesn’t have to be perfect, even a small poorly applied weld will probably be strong enough and once it’s ground smooth it’s hard to tell it wasn’t perfect! If you look around at other successful builds you’ll find a good many of them build the machine bed and base out of steel. This is because it’s cheap, quick and gives the appropriate levels of stiffness.

How big should I make my machine?

Having addressed what you should build the machine out of based on what you want to cut (I predict you chose aluminium) the next question is how big a piece of work should your machine be able to handle. Time for another prediction… you’ll start out with fairly modest aims of making a machine that can, for example, make a house number sign and the machine will quickly grow as you think of more and more projects. Try to keep control of your imagination as the larger the machine gets the more expensive it gets and the more difficult it gets to build and none of those things scale linearly!

There really is no upper limit to the size of CNC machine you could build but there is certainly a limit to the size I think you should build the first time around. If you get carried away and decide, for example, that you want your first machine build to be able to mill a full 2400x1200mm (8’x4′) sheet  of material you will probably fail or at least if you succeed it will have cost a fortune.  As the size goes up the machine has to be made stiffer and stronger to cope with the stresses of cutting and accuracy has to be maintained over a greater distance. I can think of only a handful of DIY builds that tackle machines of that size and they were all by experienced engineers.

For your first machine I would suggest sticking to a maximum cutting size of 1000mm or less on X and Y and 250mm or less on Z. I say cutting size because the foot print of the machine will be significantly larger than than the cutting area. On a 1000x1000mm machine expect the foot print to be at least 1300x1300mm and probably more. If you can keep your imagination under control I suggest keeping well under these limits if you can, every mm less is a cost saving and an improvement in the ease of building! Saying that there is an exception to the rule. If you are going with the standard three axis design with a horizontal fixed bed and moving gantry you can make the X axis up to roughly 1500mm long without too many problems or sky rocketing cost. This is because most of the X axis isn’t moving so making it longer isn’t much of a problem for the machine (the limit is caused by the ball screw and without getting into detail here it gets expensive over 1500mm).

How much will my machine cost?

This question is almost impossible to answer because there are so many variables but you have to be realistic about how much the build will cost or you will be very disappointed with the results. As a very rough rule of thumb a machine that will cut aluminium reasonably well and has a working area under the limits given above will cost somewhere between £2000 and £3000 (in 2013). It can be done for less if you buy most parts second hand and go for cheap parts but you’ve got to be careful about where you reduce the costs or you run the risk of ending up with an unreliable mess of a machine.

A machine that is destined to only cut wood can probably be built for around £1500 perhaps a little less since it could be built mainly from MDF. If you want to go very low cost you can go down the roller blade bearing route (where you make your own guides) which can dramatically reduce the cost but at the expense of additional construction work, lower reliability and greater inaccuracy. I would expect you could build a machine for under £1000 going down this route.

Personally I believe the first time builder should save up until they can afford to spend around £2000 on the machine. Spending less often leads to a machine that is compromised and or inaccurate which results in disappointment. A side benefit of spending a little more the first time around is the parts can be sold on for a good fraction of their purchase price if you decide to get rid of the machine or more likely used in a second machine built by the first machine.