Beginner in CNC trying to decide on number of axes

[doc7]

Hello, everybody.


I'm a hobbyist with a particular taste for DIY projects, customisation and general tinkering. About a year ago I bought a 3D printer and have loved it ever since - I can produce little items that I can't buy anywhere, it's cheap and pretty much imagination is the limit. I have had my taste of 3D printing and as awesome as it is it has its own limitations, mainly in the materials the process uses. I want to upgrade to a CNC router and to be able to machine items out of materials like wood and aluminium and maybe steel, but I realise that this is a whole new level.


I'm searching for a unit that will be able to machine items that I design in three dimensions. I'm willing to buy a kit but I would prefer to assemble one myself (I'm assuming it won't be too challenging for me as I have built couple of 3D printers from scratch). I did a bit of a research and found out what 3-axis routers are capable of (pretty much 2D items and quite simple 3D ones) and I don't think this will be enough for my needs. I also looked into 4-axes and 5-axes routers and I believe that this will cover what I want to do. I, however, have doubts that 4-axes will be sufficient and that 5-axes will be too elaborate of a project. I realise that a small amount of after processing will be required but this won't be a problem.


I'd appreciate if someone more knowledgeable was to clarify things a bit for me.


Best

Similar Threads:

• Need help - two slave axes of two main axes?
• Newbie- help me decide!!
• Newbie- Help to decide what to buy..
• Newbie- Just trying to decide
• Help me decide on what CNC i need

[ger21]

It really depends on what you want to make. You can actually make complex 3D shapes with a 3 axis machine, by machining the object from multiple sides.
A 4th, rotary axis basically automates this process.

I would not even consider a 5 axis machine. There's a reason that you'll find only a handful of 5 axis builds on this forum, which has been around for 14 years.

[doc7]

Thank you for your reply Gerry.

Could you go into some more detail about the differences between 3-axes and 4-axes setups. I did imagine rotating the object around with a 3-axes router but I fear that I won't be able to rotate it perfectly enough so the two (or more) sides of the object line up as intended. I like the sound of automation of that by a 4-axes router but what's the process of it? I'm guessing one just inputs the CAD file and the router does its job?

I'm shooting for versatility, so I guess 4-axes is the way to go, eh? Based on your reply I'm taking 5-axes out of the equation.

[awerby]

For materials up to aluminum in hardness, a router will work, if you build it stoutly enough. Putting a 3-phase spindle on it will help. But for steel, you really need a mill, which is a different sort of machine altogether. Routers are fairly easy to build, but mills are not; if you need one but can't afford a new one, it's usually best to start by finding an old one and retrofitting it with a new controller.

4-axis machines usually have a rotary table set parallel to the long (X) axis, with a chuck or other holding device on the rotary and a tailstock to pick up the other end. To produce a 4-axis toolpath, you need a CAM program that's capable of that; the router itself won't translate CAD files into the G-code that's needed. Once you get your router or mill operational, I'd suggest you download DeskProto, which will do that translation and generate live 4-axis toolpaths. It will work for a month for free before timing out, but it's quite affordable for hobbyists.

[doc7]

At this point I will stick to a router as a machine and aluminium as hardest material. Thanks for making a clear distinction between the two.

What are the advantages/disadvantages of both types of spindles (3-phase and regular)?

You highlighted good frame for the router. Where can I find plans for a good machine or, alternatively, what are reputable sellers and/or kits out there?

[Ryadia]

Three axis means the spindle moves along the length of the machine - across the width of the machine and up and down. (Z axis) The forth axis is not actually attached to the machine. It consists of a small lathe chunk that is controlled by a stepper motor - that's why it gets the title of an axis. I have both a 3 axis and a 4 axis machine. One I've had for a few years (3 axis) and the other (4 axis) is the single biggest headache I have ever had the misfortune to own. There is a seriously huge learning curve to use a 4 axis router. I made a chess piece (one of them) using a 4 axis router at a 'mens-shed'. It took me 2 weeks to perfect the CAD drawing of it and 6 weeks of trial and error (1 day per week) to actually machine the horse's head. I could have carved the thing in about 3 days. I wasted my money on the 4th axis.

Any new-comer to CNC who thinks they will buy a router off Ebay and start making anything at all has rocks in their head. Me included. I'm pretty smart when it comes to tool making but not smart enough to learn a CAD/CAM program to competent stage in a year or less. Then you have to deal with the Chinese whom you bought the machine from (if it came from Ebay) they is a hell of a lot smarter at figuring out customer motivation than you or me ever will be. I was lucky. I junked the counterfeit controller card and pirated software. Then stupidly bought an "All-IN-One" controller on the basis of "The only controller you will ever need" bullshirt from the Australian (clearly not born here) designer.

That makes 2 stupid mistakes I made in the same year that I'll never make again. If you can't see the thing you are buying actually working and then put back in the box(es)... Consider it doesn't work. The days of believing a seller telling you something like "Yes, I (or we) can certainly help you. I get 2 or 3 calls a week form people just like you. I wonder now if those who believed him actually returned his controller and got a refund or threw it under the bench with whatever else they bought that didn't work and took up fishing. Starting an outboard motor is a hell of lot easier than dealing with someone stupid enough to include "design features in their product that guarantees it won't work. Hows this for imaginative designing...

First thing is Nar... No one needs a mouse. We'll just set it up for a keyboard and work out some secret (as in never published) keystrokes to control it. Second thing... Hey! Ebay has created a ready market for my controller. I'll aim it at replacing the Chinese junk. Yeah. Lets add a design feature or 2 to make it look good. First feature. When you install it we'll save $2.10 (the cost of a transformer) by having it run off the Chinese switching power supply. Second feature, when you boot the controller we'll have it lock itself to avoid our stupid customers cutting off their fingers or hand. We'll make it so you have to press the Emergency Stop Button to free it up so you can start working. What do think? Good ideas or not?

Terrible ideas. The Emergency stop button does 2 things. Halts the machine and kills the power. Haven't got it yet? Go back a few lines and you'll find the answer. To start the controller in a usable state, you have to hit the Emergency stop button. That kills the power and switches off the controller. Start again and the same thing happens. This wouldn't happen if he'd have supplied the $800 controller with its own $2.10 power supply. Maybe it would have worked, maybe not. At week 2 and 9 Emails to the developer - some asking for keystroke list, some asking for instructions that I stupidly expected to be included with the controller and an hour and a half in a 3 way phone conversation to pass control of the spindle speed to the controller, I spat the dummy, stamped my feet on the ground and demanded a refund.

The sad fact is that if it was just me, I'd probably take the fishing option but every month a container full of routers arrives in Sydney for 'delivery from Australia' From a bit of research it seems that there are 10 identical sellers of these routers at any one time. They can be identified by looking at what else they sell. Baby clothes seems to be the flavor of this month but it can change from meat mincers to capsule filling machines. Ask yourself why isn't the owner of the shipping container doing the selling? Too many negative feedbacks. Whenever one of these diaper sellers gets negative feed back, 2 more pop up to take their place. And what say Ebay? nothing! No reply to email, no one to answer calls. They are as guilty as the sellers refusing to reply when you ask the magic words. PIRATE SOFTWARE or COUNTERFEIT CONTROLLER CARDS. Better idea is to pay more and got more. Get at least something that works.

Ryadia

[doc7]

Thanks for the long post, Ryadia. I'm really not leaning towards buying stuff from ebay exactly because of the things you have mentioned. The way I see is either buying a kit from a reputable source or building one myself, given someone else's schematics.

[Ryadia]

Just be very wary of the timber routers being offered in Google ads. I don't know if you get the same Ebay stuff as Australians do but you can buy a frame kit - they are quite robust from the mob selling the complete router with Pirate software and counterfeit control boards. In hindsight that's what I should have done. You can buy a kit of Gecko stuff and stepper motors from Gecko and make your own controller. I've no idea the cost relationship of doing this but it sure looks like a fine way to get a really good router.

Best of luck whichever way you go.
Ryadia

[A_Camera]

Hello, everybody.


I'm a hobbyist with a particular taste for DIY projects, customisation and general tinkering. About a year ago I bought a 3D printer and have loved it ever since - I can produce little items that I can't buy anywhere, it's cheap and pretty much imagination is the limit. I have had my taste of 3D printing and as awesome as it is it has its own limitations, mainly in the materials the process uses. I want to upgrade to a CNC router and to be able to machine items out of materials like wood and aluminium and maybe steel, but I realise that this is a whole new level.

Wood, plastic, copper and aluminium or other similar soft material is OK, but I'd not even dream about steel unless I had a very robust machine. My DIY CNC weights about 80-90kg now, has a fairly small work table area and I think it is very good and strong but I will not try steel with it because I know it is not good enough for that. For steel the mass and the rigidity must be increased several times...

I'm searching for a unit that will be able to machine items that I design in three dimensions. I'm willing to buy a kit but I would prefer to assemble one myself (I'm assuming it won't be too challenging for me as I have built couple of 3D printers from scratch). I did a bit of a research and found out what 3-axis routers are capable of (pretty much 2D items and quite simple 3D ones) and I don't think this will be enough for my needs. I also looked into 4-axes and 5-axes routers and I believe that this will cover what I want to do. I, however, have doubts that 4-axes will be sufficient and that 5-axes will be too elaborate of a project. I realise that a small amount of after processing will be required but this won't be a problem.


I'd appreciate if someone more knowledgeable was to clarify things a bit for me.


Best

I can't help you out with kits or any useful advice on what/where to buy since my CNC is 100% DIY, and is totally my own design. What I can say about buying is:

1. Don't buy anything without ball screws on all three axes.
2. Don't buy anything with too low Z clearance. I have about 160mm Z clearance and that allows me to use quality vice and long tools. This is very important. Most cheap CNC have very low Z clearance, so be careful.
3. You can make 3-D objects with a 3 axis CNC also, but need to turn around the piece you work with. I am making small, precision 3-D objects and turn them around 4 times before I am ready. A rotational axis would save me some manual turning around, but not all. So, it depends on what you want to do and how you are working. Without accurate vice and high precision in my machine I would not be able to make what I am making. Anyway, if I ever buy a rotational axis I have no problems with the clearance, 100mm rotational axis will still give me some tool clearance.
4. I would not buy a 5 axis machine. That's overkill, and I think it requires experts to use it and to build it. I'd leave that to professionals.
5. A precision vice and accurately squared 3-axis machine will give you more pleasure than a sloppy made 4 or 5 axis machine. Spend your money on a precision vice and the machine quality, not on the number of axes.

[doc7]

Thanks to Ryadia and A_Camera for the suggestions.

I have couple of questions.

What do you refer to when you say "controller".

What is the classification of materials in terms of hardness? I would go wood->plastic->aluminium but there surely are others.

[A_Camera]

Thanks to Ryadia and A_Camera for the suggestions.

I have couple of questions.

What do you refer to when you say "controller".

Controller hardware for me is the motion controller, which in my case is this:

CNCdrive - motion controls

Controller software in my case is this:

CNCdrive - motion controls

But of course, there are alternatives, like the very common Mach3 and LinuxCNC.

What is the classification of materials in terms of hardness? I would go wood->plastic->aluminium but there surely are others.

Hardness is measured in Brinell. Wood is not automatically soft, some are in fact as hard or even harder than aluminium, so you can't really say that the order you list them is the order of their hardness. There are also different plastics and aluminium, some are soft others are hard. However, most better made DIY machines can handle wood, plastic, aluminium, brass and copper. Of course, even other materials can be milled, like fibre glass, PCB, corian and so on, but I listed wood, plastics,aluminium and copper only as a guideline, not as an order of hardness.

[doc7]

Wow, I would never guess that glass is rated to be harder than steel!

About CNCdrive. They seem to be located in Hungary and, based on your position on them, I can infer they are at least decent. What are your thoughts on the company as a whole?

Also, what are other well-known manufacturers of boards and software?

[ger21]

CNC Drive is excellent. Good products, and outstanding support. I found an obscure bug, and submitted it to them Sunday morning. They initially couldn't duplicate the issue on any of their machines, but was still able to get it resolved by Monday afternoon.


There's a large variety of control software, and hardware available.
Mach3/Mach4 are the only control software programs that rely on 3rd party hardware. Mach3 can also run from a parallel port, with no hardware controller required. But this limits you to a 32 bit version of Windows 7 or older.
There are many hardware manufacturers offering Mach3 compatible boards, from cheap chines boards, to professional grade hardware.
Mach3 is by far the most popular, but I don't recommend it, as development stopped about 5 years ago. And because hardware and firmware is developed by 3rd parties, you often find that functions wwork differently, depending on the hardware your using.
Mach4 has been in development for over 5 years, with no end in sight.

LinuxCNC can also use the parallel port, and uses 3rd party hardware controllers from Mesa.

All other controls available are by companies that develop both the hardware and software.

Here are the three most common. All offere a variety of hardware options, at different proce ranges.

UCCNC by CNC Drive.
Planet CNC.
Eding CNC.

[A_Camera]

Wow, I would never guess that glass is rated to be harder than steel!

Yes. That's why you need diamond drills or similar to drill in glass but can use ordinary, standard twist drills when drilling in steel.

About CNCdrive. They seem to be located in Hungary and, based on your position on them, I can infer they are at least decent. What are your thoughts on the company as a whole?

Also, what are other well-known manufacturers of boards and software?

CNC Drive has excellent support and in my experience very good products. I don't have any other positive experience with other motion controllers, but as far as I know, ESS is also very popular and good, but is more expensive and if you use ESS you also need something else than UCCNC. I have also used Mach3 but it is a bit outdated and there is no support, other than what their users provide. Mach3 is also much more expensive than UCCNC.

[A_Camera]

CNC Drive is excellent. Good products, and outstanding support. I found an obscure bug, and submitted it to them Sunday morning. They initially couldn't duplicate the issue on any of their machines, but was still able to get it resolved by Monday afternoon.

...hmm.... occasionally they are so fast that some times I wonder if half of Hungary is working for them.

[ger21]

It also seems like they don't sleep.

[doc7]

You guys are awesome.

I only have one request to make. Could you help me configure a setup in terms of price which is +/- 10% of what it would cost in the end. In this price I want to include the body of the machine, all the motion control hardware that I'll need (motors, power supply, controller board etc), the motion control software that I'll need, the spindle and the cutting attachments, assuming I'm not forgetting anything of importance.
This will be incredibly useful for me as the price will dictate what I'll need to do with my finances in the next 1-3 months.

I'm looking for a kit with the following characteristics:
-0.5m*0.3m*0.3m to 1.5m*1m*0.5m work area
-3-axis setup
-slotted floor/base to attach vices/fastening equipment

A rough estimation will suffice but, nevertheless, it will be crucial in terms of future budgeting.

[doc7]

I did a bit of research

The way I see it goes in these two directions
1)
Openbuilds OX 1500x1000 + Electronic kit - 1 200 eur
Spingle - ?
CNCdrive Software - 50 eur

2)
Openbuilds OX 1500x1000 - 800 eur
Spindle - ?
CNCdrive Software - 50 eur
CNCdrive motion control hardware (motors, control board, power supply etc) - ?

Edit: Note that the Openbuilds OX is just an example. I saw it and I liked the concept as I have used Openbuilds before and I love it. For this phase of my search (calculating budget) this is enough, but if you recommend another platform which will make the estimation more precise I'm all ears!

[ger21]

Could you help me configure a setup in terms of price which is +/- 10% of what it would cost in the end.

No, not really.

An Openbuilds Ox, at the low end of the spectrum, is only marginally adequate for cutting wood, imo. I would not recommend one. You can find a lot of Ox users who immediately start trying to upgrade it to a better machine, or move to a different, better machine altogether,

Price could run anywhere from 1000 euro to 5000 euro. Depending on how good of a machine you want.

The issue that always comes up is expectations. Often, the people asking the questions don't even know what their expectation are, so it can be very difficult to make recommendations.

[doc7]

Let's go at it like that, then.

Could you give me some pointers where to look for an alternative to the Ox. Or if you could direct my attention to certain producers/kits, that will give me a starting point too.

What are the features a good spindle ought to have?

This should give me some food for thought.