[zild1221]

Well, after about a year of putting it off... I am finally ready to tackle my first CNC router. (I put some money aside to just do it) I have been endlessly looking at prebuilt systems (in the $1000-$1500 range), as well as DIY builds. My mind is kind of boggled on what design I want to go with. As for what I want out of the machine:

-2'x2' cutting area

-Ability to work with woods, plastics, and aluminum.

-Fairly accurate. I'm not making jets or anything... I don't know what really to expect accuracy wise. I want to be able to use this machine to make parts for another larger machine in the future.

My problems... I don't have any tools to machine metals. I have a drill press I got from an auction, bench grinder, the ULTRA basics. I do however have access to every woodworking tool imaginable.

The goods... My family owns a hardware store. So I can get things like the router, any fasteners, and other parts at wholesale.


Taking all of the above into account, this is what I am thinking.

-MDF & aluminum extrusion structure

-Fixed Gantry (I want it to be SOLID for aluminum work, even if it means a larger footprint)

-No vbearings or skates or anything of that nature. I want to incorporate full rails with block bearings. With the size being 2'x2', the rails are going to need to be supported. I was thinking of doing this http://n0m1.com/2012/04/23/on-the-ch...d-linear-rail/ . Would be a good way to learn to tap as well.

-I don't think for what I want to do I need ball screws. Acme Precision Lead Screws should suffice. I don't know what start or pitch I should be getting though.

-Driver HobbyCNC PRO Chopper Driver Board Kit | HobbyCNC

-Motors not sure about yet.

I'm working on a rough sketchup model as I type, so the above may seem kind of scattered. Let me know what you guys think and what I should adjust before I go further.

[harryn]

Rule of thumb - your machine frame should be made from a material harder than what you plan to machine. MDF is not harder than Aluminum, but steel is.

Don't be afraid to make a proto from plywood, and then get help from someone to machine the final steel parts. The proto parts and all of the changes are what really cost money, not getting the final part made.

I have found that the local craigslist has adv from local people that will do side machining jobs, and some adv that they will help you build hobby projects on their manual mill. There are more mills out there than you might initially think.

If you make your framing material from 1/4 in thick stock, you can tap it, which can be very handy sometimes.

Good Luck and have fun.

[louieatienza]

Glacern Machine Tools - Linear Rails and Bearings

As for material... I've successfully done many jobs cutting aluminum, using a machine I made of 1" phenolic. I did however incorporate used linear rails I bought off eBay. The phenolic is a bot more impervious to climate than MDF, you can cut it with woodworking tools, it's very strong and stiff, and can be cheaper than aluminum plate.

[TechGuy5002]

I would recommend holding off on trying to build a CNC and purchase a manual vertical mill first:

1. It best to become familiar with manual machining so you can practice the basics. Any tooling you purchase (endmills, vises, step clamps, etc) can also be used with a CNC mill.
2. The manual mill will permit to you fabricate precision parts for your CNC. This will provide the means to square off the ends of sawed parts (from either a chopsaw or even hacksaw cut parts) It will also provide the means to machine parts with exact lengths (by maching two or more parts together so they have the exact same length).
3. I would recommend a reasonable sized mill that can has a work space of 20"X7"X14". Need good Z clearance for tooling (rotary table, vise, etc)

I can guarantee that you will not be disappointed owning a manual mill. You will find lots of uses, even if you have a CNC machine. A reasonable mill will cost around $1500 to 2K (new). You may be able to purchase a used mill for less. Also consider the tooling cost, you will probably spend another $1K to $2K on tooling (endmills, vises, step-clamps, collets, rotary table, cutting fluids, etc). Although tooling can be purchased over a period as you need them.


For milling alum. parts using a 2'X2' work area, the machine will need to be made out of heavy steel, I suppose stiff alum like 7075 would work, but it would be cheaper to use steel. Using soft alum or non metal materials will permit too much deflection causing unbelievable tool chatter and a poor finish in alum.

For any precision, your DIY CNC machine will require precision ballscrews. These will cost more than $1500 for a set of three (X,Y,Z), or you will need to use a set of linear encoders with a cheap set of ballscrews or leadscrews. but the cost of the encoders and cheap screws will probably be just as costly, as precision ball screws.

FWIW: a good DIY CNC machine with a 2'X2'X12' work area with precision ballscrews with a steel frame will probably cost between $5K and $8K.

[louieatienza]

While steel may be obviously stiffer than aluminum, it also weighs three times as much for the same size piece. This means heavier duty everything, from linear rails, to drive motors, larger screws, etc. It's also going to require heavier duty machinery to fabricate. All of this is going to add up to the cost. Plus, steel may not be the best material if ridgidity and vibration are a concern; cast iron would probably be a better choice.

You don't absolutely need precision ground ballscrews; most manual mills and many CNC mill still use ACME screws.

Aluminum is pretty mch free-cutting, and it doesn't take as much force as some would lead to believe. You can work around the deflection and still get an accurate part, if you do rough and finish passes, and use the right parameters. Sure more ridgidity is better but it always comes at a cost, which is the determining factor here.

The other thing is to assess exactly what you will be using your machine for. CAM and CAD software can get expensive. So are toolong. There are many killer builds here, where the builder spends a fortune on steel this, and linear that, and servos and on and on, then go out anf buy the cheapest POS bits and use basic CAM software and never fully reap teh potential of their machine. Then there are those who "theorize" about stiffness and FEA and such with no CNC machine to show for it. Heck, if I had my way, my machine would be made entirely of cast iron, with all bearing surfaces ground and hardened, with all the bells and whistles like servos, ground ballscrews, etc. But you have to build within your means, or wait (possibly until the fifth of never) to get the "dream" parts....

[adprinter]

My problems... I don't have any tools to machine metals. I have a drill press I got from an auction, bench grinder, the ULTRA basics. I do however have access to every woodworking tool imaginable.

The goods... My family owns a hardware store. So I can get things like the router, any fasteners, and other parts at wholesale.

If you are truly bent on starting down this path, I would recommend that you first consentrate on tooling up! Metal working, at a minimum, requires access to a machine shop for the basics, like cutting metal. A horizontal-vertical metal-cutting bandsaw is almost an essential basic.

You mentioned a drill press. This is also an essential tool, but will serve you better if you also invest in a milling table (Shars, has a good one for about $200). As hole locations are difficult to achieve without such a system. Metal working is a very different animal from woodworking.

Precise hole locations for items to be bolted together are absolutely required, when working with metals. Unlike wood, which can be "persuaded" by use of a hammer! If you first obtain these tools, there is a lot you can do yourself along the way. But without them, (or access to a shop that already has these tools) you will find yourself at the mercy of time and $$$ in every little thing you set about building!

Machine building is a very time-consuming process. And you can achieve very good results, by taking the time to first "think things through". But you will learn, that no matter how much you try to do this, there will always be some point in your build- when you will still end up changing the original design, right in the middle of things. Don't worry about it, because it will happen!

The main thing is, to just get started. Staring at a computer screen, just doesn't teach the same lessons as hands-on an actual build will do. Welcome to the addiction of the DIY world of CNC!

[DonFrambach]

Have you considered either buying a kit or at least parts from CNCRouterParts.com? I made a 14"x14"x5" fixed gantry router from aluminum extrusions and parts from CNCRouterParts.com and Glacern.com. My router is stiff enough to mill aluminum and brass. A table saw with a cross-cutting sled, a drill press, and hand tools would have been sufficient for my build. I did use a small CNC mill for a couple of parts but those parts weren't essential for the build.

Whatever you decide to do, I hope you make a build log. I really enjoy seeing and learning from how others build their machines!

[roy_okc]

@Zild1221,

I built my 80/20 rig without metal working tools other than a drill press and cross-slide vise. Aluminum can be cut with a carbide blade in a circular saw or better yet a miter saw. The (Harbor Freight) cross-slide vise was extremely valuable in getting accurate enough and especially repeatable holes.

While I have no personal experience yet milling aluminum I've read plenty of posts and seen enough Youtube videos of people doing it with wood and aluminum framed machines. The caveat with these is that you have to go slow. So you need to determine how much aluminum you really need to mill. If you expect to do daily, then a mill is probably the better way to go; if once or month or less and high accuracy not needed as you stated, then a router may suffice.

Try to find used linear bearings on eBay or perhaps even better a local tech salvage place. You might luck into something even cheaper than buying new vbearing/skate setups would run.

If you go with 80/20 extrusion, CNCRouterParts has some things that will make your build go MUCH easier. I would have had a very difficult time building equivalents with my tool inventory.

I'd highly suggest looking at a Gecko driver, probably the G540. Yes, it costs more than most, but read some threads of people who started with something else, had problems, and eventually moved to a Gecko. Look at Keling's offerings for driver, stepper, and power supply options. His kit #1 at Page Title is probably about the most efficient setup for a small to medium rig (you need to research all the electrical specs involved to see that the 381 oz-in @ 48v on the G540 has more power than higher oz-in steppers). Crevice Reamer on CNCZone has a wonderful write-up on his own web page on calculating the motor specs vs power supply vs driver specs, etc., unfortunately, I can't find it at the moment.

Since you aren't expecting to need super accuracy, I'd recommend 1/2"-10 TPI 5-start Acme. This will still give you very good accuracy assuming that everything is tight and you have a decent anti-backlash nut (again, see CNCRouterParts for easy and decently priced solutions). You'll also have good speed with 5-start. For instance, with 1/2"-10 5-TPI rod, typical stepper w/200 steps/turn, and G540 10 microstepping, you'd get a theoretical 0.00025" precision (theoretical because these machines + microstepping can't hold at such tolerances). With a decently tight machine and good bits and good procedures, though, you could expect smooth 2D and 2.5D results and accuracy and repeatability with a few thousandths.