Getting Started BUilding a CNC Machine

Hi,

I have done a good amount of 3d printing, and it is getting kind of boring. I want to build a CNC mill, because I like building not buying, but I am kind of lost. I know I need stepper motors, linear rails, leads screw, etc., but I am confused about the electronics + coding. Can someone tell me how all the electronics and coding work. Is it Arduino that sending signals to stepper drivers, then steppers, or is there a board with drivers built in were the "code" goes to? I want to use the best quality I can, so if Arduino is not good enough, can someone tell me what is? Does the code go straight from Mach 4 to the board or is there another code in the board( beside Mach 4) that accepts the Gcode from Mach 4. If so, what code? Also, do tool paths from Fusion 360 go to Mach 4 then said board? I would be great if someone could possible lay everything out. Also, what spindle(not crazy expensive) should I get for mostly aluminum, but some harder materials. I do not want to have to use coolant for anything. Also I want the spindle that accepts ER collets with that quick release mechanism. Thanks

Hi James,
Forgetting about ardunios, mach4 etc it goes like this:

1) You design something in CAD, this could be a 2D thing or a 3D thing
2) You export the CAD into a CAM (computer added manufacture) program. The CAD and CAM could be integrated and not separated eg rhino3D CAD and Rhino3d CAM are integrated like Fusion360 they are inside the same environment. I use three different CAD programs and import the CAD into a free standing CAM (mechsoft mill or UCCNC) depending on what sort of job it is
3) The CAM program is used to create G code and M code. Gcode is a toolpath which is a string of 3D co-ordinates that tells the machine where to move the tool to. This path can be checked or simulated in toolpath checking software. M code are commands like turn spindle on and off, turn on lights turn on/off air blast etc
4) Once you are happy with the Gcode this is imported into a machine controller. eg mach4, UCCNC, acorn etc etc there are many controllers around
5) The machine controller turns the Gcode or NC code into machine impulses and co-ordinates the axes, it sends out a stream of impulses to each motor in the system
6) But these impulses are also feed through a motion controller. This looks ahead and identifies corners and stops and controls the velocity of the motors. If there was no motion controller (like a UC100 that I use) the machine would crash as it doesn't know that a corner is coming and needs to slow down
7) Once all of this is done the tool moves at the speed it should, the path it should and you get a result.

Cheers Peter

Regarding other requests 1) making a mill to cut aluminium is a tough call for a newbie plus cutting dry is fraught with difficulties. Do lots of reading in the forum about this requirement. You do say a mill so it is mainly for metalwork 2) ER collets are not quick release? The tool has to be set every time. 3) look at Taig mills or similiar... if these are too small then look at mills of a size you need and have a good think... Mills are large chunks of metal, maybe think about converting a small manual mill to cnc then lots of the hard work is done...

Answering all those questions would be like writing a book, James. Posting them one by one would probably get you more responses. But I'll try to reply to some of them.

First off, most of the machines people build from scratch here are routers, not mills. Routers are a lot easier to build because they don't need to be as rigid. They're used for cutting foam, plastics, and wood. The most rigid and heavy routers can handle aluminum to some degree, but not steel. Milling machines are designed for cutting metals, although they can handle softer materials to some extent. Mills typically are built to deal with flood coolant; routers usually have some kind of dust collection. Mills generally have a metal T-slot table; routers mostly have wood-based spoilboards for attaching the workpiece to. Mills typically have a smaller work envelope than routers of the same price, and usually they're longer than they are wide. Routers can be much larger, but normally they have less headroom than a knee mill, which can accommodate higher workpieces. So you should identify the sort of things you want to make, how large and in what materials, to decide what sort of machine to build. If you really want a mill, not a router, you should consider buying a used one with an obsolete control and retrofitting it - this will get you a better machine than you can build yourself, unless you happen to run an iron foundry.

There are a lot of different ways to make a CNC mill run. There are proprietary controls you can buy and open-source ones you can build, for instance using Arduino chips. Obviously you need to know more to build a CNC control than just to run one, but for choosing one, you mainly need to know how much speed and torque your axis motors will require. Arduino-based stepper-motor systems are at the low end of this spectrum; proprietary systems based on AC servo motors are at the high end. So when you speak of wanting the "best quality" electronics, that needs to be put in context. Putting an Arduino-based motion control system on a high-powered VMC wouldn't make sense, nor would putting a Centroid control on a router made from MDF and skateboard wheels.

The normal flow of operations for a modern CNC operation involves first creating a model in a CAD program representing the part one wishes to produce. This is then fed into a CAM program (such as Fusion 360), which generates one or more tool paths for the various tools necessary to carve out the part from a block of material. This is "post-processed" in the CAM software, which translates the toolpath into G-code specifically tailored to the control software (also called "G-code interpreter") which runs your particular machine. Mach4 is an example of CNC control software; it generates a series of pulses based on a protocol called "step and direction" which are communicated to the drivers that actually run the motors that move the axes of the machine, as well as regulating the spindle and whatever auxiliary functions are also available.

Spindles used in routers are typically high-speed, low-torque devices suitable for cutting materials up to aluminum in hardness. Mill spindles are the opposite - they tend to have high torque at low speeds but don't go very fast. They are able to cut aluminum, but are designed for steel and metals of similar hardness. All these metals cut much better with flood cooling, which also has the benefit of blasting chips out of the cut, so the tool doesn't get fouled by re cutting them. So by abjuring the use of coolant, you're severely limiting the functionality of any mill you get.

First let's define the machine

The following are only examples of machine type, not recommendations.

This is a table top CNC Milling Machine, suitable for metal and can be used for wood and plastic. https://www.microkinetics.com/index....xpress/express

This is a table top CNC Router, suitable for wood and plastic, maybe some light milling aluminum and other soft materials. idea – insedia.me

Note the construction differences. Either type of machine is available from many vendors and qualities. It is possible to build a router style machine that is suitable for metal, but the construction needs to be up to the task and very heavy. In general for metal, the maximum spindle speed needs to be about 1/2 of the router spindle speed (10,000 rpm), and needs to be able to operate in the <1000 RPM range. The generally available import router spindles will not do this.

All of the CNC controllers take G code generated by a CAM program and translate that into position data to run the axis motors. The position data is in step or encoder pulses depending on the exact system. Exactly how this is done and the internal code used depends on the system, every vendor has their own ideas about how things are done.

Centroid Acorn is an inexpensive and reasonably robust system as are the Dynomotion Kflop/Kanalog products.

Hi,
I built a small (machining volume 180mm x 180mm x 180mm) mill, its great. I can do engraving with it with a highspeed low torque spindle and ER11 collets for
things like making circuit boards. The highspeed spindle does a good job in aluminum with tools 8-10mm or less in diameter. What the highspeed spindle will
no do, or at least practically, is mill steel.

For steel I made a second spindle that does 3500 rpm and 6.2NM (cont) torque.

I can only reiterate what has already been said, if you want to mill steel that is a WHOLE different ball game than a router machining wood, plastics, and with care
aluminum.

The rigidity of a machine to mill steel is something like 10 times higher than a machine to mill wood and as a consequence the price to build such a machine
goes up exponentially.

A cheap 24000 rpm spindle and VFD from China might cost $300. IF you want a quick change setup, also called A(utomatic) T(ool) C(hange) then it will cost more
like $1300 for the same (cheap) quality.

I have just priced a German made (quality) 42000 rpm 2.2kW spindle with an HSK ATC and the spindle motor alone is 5400 Euro!!!! As much as I want one
I don't have that sort of money.

This rather illustrates the point that you NEED to do your homework BEFORE you commit to building anything or your budget will blow out big time and the project
will never get finished.

By the way, I use Mach4 with an Ethernet SmoothStepper (motion control board) and it works really well. Others have reported that Centroid is good value and yet others
have reported that UCCNC is very good. None of these combinations are free but approx $500-$1000 would get a really good system (from any of these three types) that
would not be put to shame by industrial controllers costing $20,000 plus.

Craig

Thank you Joe. I definitely want to build a mill. I think I will just stick to aluminum though.

- - - Updated - - -

Thank you Jim. I will probably build a mill strictly for aluminum.

Thank you for the clarification. I am definitely thinking mill.

I just looked at the Taig mini mill. That might be a better option. How is the quality?

Hi,
the quality is commensurate with the price. There are many happy users out there, it does require you being realistic about
the capabilities of such a machine. It would be a great starting point.

As someone posted retrofitting an existing manual mill might be a good option. The are plenty of 'drill mills' out there and if you
are patient you'll pick one up cheaply eventually. The retrofit would probably satisfy you 'I want to build it' urge and yet you can
use a manufacturers cast iron.....and cast iron is still the standard for rigid machines.

You might have heard that every man and his dog has jumped on the epoxy granite solution. There have been some excellent examples made,
but epoxy granite is not cheap, either in terms of materials or techniques/skills in molding and casting epoxy.

As good as epoxy granite is do you see machines made by Hass, or Okuma, or Mori Seiki made with it? No. Does that tell you what their design
engineers think?

For an aluminum capable machine I think a retro fitted drill mill would be good project which if approached sensibly could provide a capable machine
for a modest budget. Of course if at a later stage you want to replace the original spindle with an ATC spindle, or replace steppers with servos,
or replace rolled ballscrews with ground ballscrews you can. Ultimately the capability of the machine is less about what you bolt on but the rigidity
of the frame and axis beds. The more rigid they are the more satisfaction and scope it will provide down the track.

We have a drill mill at work, its used mainly as a drill press these days as we have a much bigger knee mill, but its great for small jobs including
steel if approached with care.

Craig

Hello Craig - I think the logic that because some well known machine builders use cast iron then its the "standard" (as in best) is poor logic. There are many, many machine builders using other materials. A quick search has shown these up using variations on epoxy granite.

https://www.rampf-group.com/en-cn/pr...ds-components/

https://www.basetek.com/

https://www.castinite.com/machine-casting/

https://www.microplan-group.com/en/p...e-granite.html

https://www.studer.com/en/

Times are changing and materials are changing. The companies you mention no doubt have material developments in the works to solve future machine designs. Aircraft, cars and anything that moves have gone from steel to aluminium to composites in my lifetime, so it's about time machine tools did the same. Machine builders do not use materials in their advertising so we know little about what they use. This is because materials do not sell machines, their performance and cost does.

and your mentioned Mori Seiki do make mineral composite machine parts. See attached snip from their website. Times are changing. Machine builders that want to make machines that move at 2g or more have to use lighter and damper materials then cast iron. Epoxy granite is just one of the candidates for this application.... Peter

Hi Peter,
I did agree that some very credible designs have been built and executed in epoxy granite.

This thread however is about a person looking to make a first step into CNC. For that purpose a cast iron drill mill, if one can be had for cheap,
as a retrofit would be an excellent place to start. Cast iron and dovetails are well known and still widely used......a very cost effective way into
CNC.

Craig

Hi Craig - I did say early on that a suitable approach is to retro a small mill for this project or buy a Taig. Its up to James how big an Elephant he wants to build and to what technology he uses and what his capabilities are or how far he wants to be extended. The forum is here to help. By the way James welcome to the asylum :) Peter

Hi,

Quote:

---

I did say early on that a suitable approach is to retro a small mill for this project or buy a Taig.

---

Yes you did, and very sage advice too.

When I first started I did not enough research and/or underestimated the problems and scale of what I wanted. I did learn...eventually.
I should have started as you have recommended to James, I might have made more progress or certainly achieved more with my money
than I have done.

Looking back the single most telling decision in terms of design and build consequences were that my machine should mill steel and stainless.
That one decision rather sets the tone and determines the costs of the project. Had I been somewhat less ambitious I would have saved quite
a few dollars. None the less I am happy with where I am at.

As you may recall I am in the process of a new mill build, and just today have confirmed the order to cast (in grey iron) the beds for the project.
I had hoped to have them cast before Christmas but things just got too hectic but they are underway now.

I too wish James the best of luck, if a hobby is measrured by what you have to learn in the pursuit of it .......hobby CNC must rate very
highly indeed.

Craig

Hi Craig - I think we are the sum total of our mistakes not out successes. Successes come and go but we do not learn much from them. Failures however sharpen our focus and burn themselves into our psyche. Being a senior machine designer I can recall some very commercial and technical failures that at the time were emotionally catastrophic. But given some time to heal that experience pays off. Someone will say lets do this!! Sounds like a great idea and you go hey have a deep think about that and watch out for XYZ thats the hidden sting and they will later say yeah hadn't thought about that, thanks for the heads up. You just saved them the $10,000 oops! and you can increase your pay rate...Peter

Thank you for the help. I will probably just buy a Taig mini mill and play around with that.

Quote:

---

Originally Posted by James123456789

Thank you for the help. I will probably just buy a Taig mini mill and play around with that.

---

If that's what you want, I can help with that; I'm a distributor for Taig and can offer a discount of 10% off their list prices on any of their products.

Maybe, can you do any better. What are the differences in the CNC mini mills. Their are a lot of different ones.

What is the full build volume on the Taig mini mill. Tolerances?

Hi James - Taig fully spec their machines on their site https://taigtools.com Peter

James. This is a trade or hobby with a lot of moving parts. Reading about where to go or what to do with a new machine is daunting.

I see you refer to tolerances. What do you know about tolerances? You can or might learn to compensate with a tool path change or fudging a tool diameter, or maybe climb vs conventional cut to accomplish said tolerances.

Most consumer mills and lathes, Sherline, Taig, Grizlly, PM Matthews etc all build machines that will blow the accuracy of a 3D printer away. If you listen here or on sites similar, you will hear expectations from a 1000000.00 dollar machine from a 1500.00 dollar machine. It isnt going to happen. If you are looking at accuracy or tolerance of .001" or less, forget it. Then get out your feeler gauges, look at a .001' blade and tell me your part wouldnt fit. If it doesnt, your design is done incorrectly. Get started more than worry about aerospace quality.