Thoughts on a beginner CNC

I want to start off by saying I'm not marketing anything, I'm just working on a beginner CNC project for Uni and wanted some feedback. It's about a budget orientated (sorry TDD, no linear rails) CNC router that'll be able to do woods, plastics and some aluminium. So if you're wanting to get into CNC machining, and you're on a budget, boy do I have some questions for you! Without further ado, here's my little survey, feel free to comment on any of the points below or anything else you might feel is relevant. Feel free to extrapolate your experiences you've had with 3D printers here, more or less the same ideas apply.

1) If you own a CNC, anything that made it particularly easy or hard to get started. This includes right from assembly, if any, to the first cuts, so CAM and all that included. If you own a 3D printer, would you mind chipping in on what made getting started with the printer a good or bad experience? Assembly instructions (or lack there-of), pre-configured slicer profiles, support from the company etc.
2) What are some features you consider essential or non essential for a first CNC? This can be anything from work holding, touch plates for automatic homing, whatever. If you don't have a CNC (but want one), what's your list of must haves?

3) For those on a budget, but wanting a CNC, what sort of budget are you looking at? There are plenty of machines like the MPCNC, R3, and other homemade CNC machine made of MDF or plywood. These are all obviously budget focused machines, and if you were looking into one of these what sort of price point would make you consider buying a kit instead of sourcing and doing it all yourself? 3D printers really took off when machines stopped costing into the thousands, so what sort of price point do you think would make CNC accessible.
4) What's the most important thing for you when looking at a CNC (or when you were looking at buying a 3D printer)? Cost? Support? Precision (what sort of tolerances are you wanting?)? Work envelop? Good instructions (for assembly, operation, CAM)?

Lastly is just what sort of machine do people want? Big, small, does it have to be super precise or is +-0.1 mm good enough? What sort of budget would you be looking at?

Thanks in advance for chipping in any opinions or thoughts, any feedback is appreciated, doesn't necessarily have to be based on the questions above.

Edited 1 time(s). Last edit at 04/12/2018 01:56AM by Trakyan.

Try posting this on cnczone.

I have an original Shapeoko. I think I paid $200ish for the kit minus electrics and motors when I bought it 4 years ago. My big purchase hobby budget is somewhere between $300-400 price tag. I have family and more important responsibilities that keep me from budgeting anymore towards hobby stuff. Important features, reliability, ease of use. CNC requires more setup when your ready to use it, so I don’t want to be fiddling with it when I’m ready to do work.

I should clarify that this is a RepRap style CNC. I'm starting with 3D printing the parts for it, then I'm hoping to CNC mill the parts for it. From what I gather RepRap isn't necessarily about 3D printing, just about using a technology to self replicate. Besides, you could always strap a hotend to it

Also, thanks for the feedback Mike, much appreciated.

Edited 1 time(s). Last edit at 04/12/2018 06:13PM by Trakyan.

Ive been a machinist for 30+ years and own one of those real vertical machining centers. My advice is if your going to cut any type of metal, rigidity in the machine and work holding is paramount. Sure you can cut aluminum without a rigid set up but it will look like poo and forget holding tight dimensions. Surface finish will suck. So whatever you decide think bulky with lots of mass for the machine structure like heavy wall square steel tube or better yet solid. look at a professional machine. Big cast iron structure. my machine weights 8000lbs. Also you typically machine materials much slower than you 3d print them like 10mm/s or less. The more rigid = faster assuming you got the horsepower. All that still applies even if your going inexpensive on the linear guides

If your used to 3d printers you will be shocked at how a) nosy cnc is b) how incredibly messy it is.

I wonder, if a MPCNC would benefit from filling up the tubes with concrete or quarzsand and resin?

Thanks for the responses guys.
@Bill Clark, I fully understand where you're coming from, but a cast iron, 8000 lb machine isn't really affordable or desktop (at least for most people), and that's what I'm going for. I'm talking more about routers than mills, things like the X-Carve, Shapeoko, R3, MPCNC and the like which can machine woods, plastics and aluminium (slowly). Making a cheap milling machine isn't realistic from my point of view, which is why I'm calling it a router. There is nothing I can do to make a "proper" router with rails, heavy duty gantry and so on that someone else can't do cheaper. So what I'm trying to do is taking an approach similar to the MPCNC and R3, trying to build something different that is inherently cheaper.

@Dust, totally with you here, but I'm not sure I can do much about that. Even engravers are pretty bloody noisy depending on what you're engraving. Depending on what sort of response I get on the machine size and cost side of things, I might switch tack from something large format to something small and enclosed.

@O_lampe, I've seen it discussed on their forums, but don't know if anyone's tried it. I assume it would dampen vibration a little. That being said it might cause more sag in the tubes and will significantly increase moving mass. I reckon finding some way to fully (or mostly) support the perimeter rails would have a bigger impact on rigidity. Then maybe doubling up on the crossing XY rails similar to what zortrax does, but then cost goes up and I lose track of my original target.

Anyone got any more opinions on the price point such as machine would need to hit? Also any thoughts on the work envelope? I'm personally leaning in favor of a large format machine because I have 1-2 projects that would use the size, but other than that I think the majority of my work would fit into a 300x300 mm work area quite comfortably.

I have built several CNC milling machines and one of these uses the mechanism designed for a 3D printer and I think the following points are worth stressing even though they have already been by Bill Clark and Dust.

• The forces involved in milling are much greater than in 3D printing. Although it is possible to machine even quite tough materials with small cutters, it will take a much (Much) longer time.
• It is, as Dust says above, very messy. In any practical RepRap type milling machine the swarf must be removed before it has a chance to get into the works. The vacuum cleaner that I use is very noisy.

Having said that, it is still possible to do some milling with a lightweight machine. Examples include isolation routing of PCBs, engraving labels in laminates and even making plaques with a limited depth. The 3D printer based milling machine I mentioned can be seen and heard on [www.youtube.com] For machining of any appreciable size even Nylon is a bit too tough to be a practical proposition although machinable wax is a possibility.

On the up-side, even though real CNC machinists will look down their noses at a milling machine such as you propose, it will make it possible to have an idea and turn it into an object in a couple of hours - and to do that on something you can move about without a crane.

Mike

Edit:

Quote
Trakyan
Anyone got any more opinions on the price point such as machine would need to hit? Also any thoughts on the work envelope? I'm personally leaning in favor of a large format machine because I have 1-2 projects that would use the size, but other than that I think the majority of my work would fit into a 300x300 mm work area quite comfortably.

No, no, a million time no. Even very limited machining to any acceptable standard becomes almost exponentialy more difficult and expensive with greater size. Go for 100mm x 100mm x 100mm and a price break somewhere near or even less than that for a 3D printer

Edited 1 time(s). Last edit at 04/13/2018 04:37AM by leadinglights.

Not long ago, the very same subject has been discussed here. Same remarks.

Now, as you are focused on making it cheap, simple, don't ask, just do it. It is cheap so not much to loose and a lot to learn, for cheap.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

@leadinglights
While I get what you mean, there are relatively affordable, large format CNCs like the MPCNC and R3. They can hold tolerances that are good enough for most people. I also understand that 3D printer mechanics aren't made for milling or routing, which is why I don't plan to use 8 mm linear rods in a large format machine, or even in a smallish format machine. I think 100x100 mm is a bit too limiting for most people, 200x200 is a bit more acceptable but I'd personally aim for 300x300.
Also, when you say a price point at or near a 3D printer, what printers do you mean? $200 rock bottom kits from china, $500-1000 machines or things like the ultimaker, lulzbot and the likes which are several thousand dollars? I want to build this machine to a price point. If $300 is what it takes, that's what I'll aim for. If $400 is acceptable, I'll squeeze in every little feature I can for that extra $100. But to do this I need to know the price point I'm working to.

Yes, this machine will be a pathetic waste of time according to any self respecting machinist, and wont be able to hold the tight tolerances they expect. On the other hand, it will cost a small fraction of their knee mill and should hold at the very least the same tolerances people get from their 3D printers. I really want to stress that this will not be a machine for people running a machine shop or looking for high quality precision parts, it's for someone who can't afford a precision machine and is happy with "good enough". I personally would be fine with +-0.1 mm, after all that's the ballpark people get from 3D printers and it seems to be just fine for most people.


@MKSA
I mentioned this is related to a Uni project, part of this is doing some market research. The project itself is already underway, I'm just working on integrating some way to square the Z axis, after that I just need to add some mounting holes to the corner assemblies and it's done. Only problem is finding a way to add adjustment points to square the Z axis has had me stumped for a week. On most CNC machines with a moving gantry like the X-Carve/Shapeoko it's fairly easy. Rotating the X axis rails about their axis squares the z axis in one way, rotating the Z axis rail about the Y axis squares it the other way. Both can be adjusted fairly easily and independently. The design I'm working on is similar to the MPCNC, where the Z axis is mounted to the intersection between the X and Y axis and I'm finding it hard to think of an elegant solution for squaring the Z axis since it's effectively bolted to two walls in a corner, you can't adjust the rotation in one axis without loosening both. From what I can tell the MPCNC achieves this by just loosening/tightening bolts until it ends up square, this just seems wrong to me.

Edited 3 time(s). Last edit at 04/13/2018 05:56AM by Trakyan.

Quote
Trakyan
Thanks for the responses guys.
@Bill Clark, I fully understand where you're coming from, but a cast iron, 8000 lb machine isn't really affordable or desktop (at least for most people), and that's what I'm going for. I'm talking more about routers than mills, things like the X-Carve, Shapeoko, R3, MPCNC and the like which can machine woods, plastics and aluminium (slowly). Making a cheap milling machine isn't realistic from my point of view, which is why I'm calling it a router. There is nothing I can do to make a "proper" router with rails, heavy duty gantry and so on that someone else can't do cheaper. So what I'm trying to do is taking an approach similar to the MPCNC and R3, trying to build something different that is inherently cheaper.

.

I got ya. my point was simply make it heavy if possible. heavy can still be cheap. The end result is you get outside the resonance. The drawback to heavy is the mass has to accel/decel but thats only a problem on a 3d printer

Heavy gets expensive for two reasons usually, more/more expensive material and higher transport/shipping costs. I like the fact the mpcnc creator has found a way around this by having the heavy parts all be easy and cheap to source locally and I plan to do the same. You're right about extra mass helping with resonance and vibration though so perhaps in the build instructions I can add something like what o_lampe suggested about filling the tubes.

Another thing is, as a machinist you might be able to help me out here since I'm guessing you have more experience with metal stock than I do. Do you think either round or square tubing is a better choice as a "rail". I know neither is intended for this use, but like I said I'm building to a price point. Have you found either to be more or less straight or in spec than the other?

Now if only government would give the same loan deal for a CNC machine as you get going to Uni.
Low interest, pay it back when you start earning...i'd take that deal...but they can shove the Uni deal where the sun dont shine
right up their over paid....

Concrete is cheap, heavy, fit for the job (time proven since 1915). The new ones; reinforced with fiber, epoxy, steel are even better.
[makezine.com]

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Quote
Trakyan

. Do you think either round or square tubing is a better choice as a "rail". I know neither is intended for this use, but like I said I'm building to a price point. Have you found either to be more or less straight or in spec than the other?

Do you mean using either as a linear slide/guide?

Yeah, I know that's not what they're meant for but they're cheap which is the aim of the game for me.

Maybe round because it can be more forgiving in certain planes for misalignment and potentially have better contact for a given area . What are you considering? pics,links ?

Edited 1 time(s). Last edit at 04/15/2018 08:39PM by Bill Clark.

At the moment the design is build for a generic round rail of any given diameter, but I can tweak it to fit a square rail fairly easily.

Nothing in particular in mind as far as what round stock/tube I plan to use. I live in New Zealand so everything is super expensive here and there isn't as much selection at our hardware stores. There is only really one option for steel tube, I think it's galvanized but it's not specified. Neither is whether it's hot or cold rolled etc. To be honest I had trouble finding square tube of a suitable size, they didn't have much bigger than 15 mm and I think it was all aluminium.

If you have a 3d printer and some nylon you could print a bushing for any combination of what ever you can get your hands on. I guess import fees are a problem where your at?

Not so much import fees as shipping, I pay more for the shipping than I do for the filament itself usually. Import fees are pretty generous here, I think they only kick in on things costing over $400. I've debated bushings, I just think that they wont ride well enough on something like galvanized steel tube/pipe. The friction would probably cause some issues, along with inconsistency in the tube diameter. I feel like a carriage with some radial ball bearings as rollers would be much more forgiving and give smoother movement. I just haven't seen anyone try or use bushings on steel pipe, so I'm guessing there is something wrong with that. I would love to be proven wrong, though, I'm a big fan of bushings and it would greatly simplify the design and part cost. I'd love to experiment but I don't think I can spare the cash at the moment.

If anyone has any experiences with bushings on hardware store piping/tubing please chip in, I'd love to hear any experiences.

When you have to pay so much for filament, you need a bulletproof design. Dealing with bushings pressed in printed parts can get frustrating ( and expensive ) pretty soon. They are always too loose or too tight. ( Murphy's law )
I have changed the design of my Compact Carriers many times until I got a version that I can print on any printer, every day. But I had to add some adjusters and clamping screws and such.

I think that's why the majority of printed bushings have some sort of 'spring' mechanism to them that lets them flex around a shaft. The issue being that the diameter inconsistency of steel tube is probably too great for this to work, and the rough surface finish doesn't agree with bushings. I might make a small version with smooth rods and bushings one day.

Right now I'm going with six radial ball bearings on a carriage, arranged in two sets of three bearings arranged 120 degrees apart. The carriage is designed to be a slight interference fit with the intent that the plastic will deform a bit to accommodate the tubing, giving a carriage that rides along the pipe with zero slop and can flex to tolerate inconsistency in the diameter. It's no precision linear guide, but those use the same concept (the carriage is a slight interference fit) to eliminate slop. I'll have to work out what the right amount of interference is through trail and error, but that's unavoidable. I might add a tension bolt that will squeeze the carriage around the tube, but I'm pretty averse to adding more parts, I'd rather just get the interference fit right than add in unnecessary parts and complexity to make up for my slight laziness.

You don't have to use the pipe as_is.
I know it's a lot of work, but could you treat the pipe surface to become smoother? Sanding it down, while it spins...
But be careful with zinc-plated iron pipe, don't inhale the dust.

Quote
o_lampe
You don't have to use the pipe as_is.
I know it's a lot of work, but could you treat the pipe surface to become smoother? Sanding it down, while it spins...
But be careful with zinc-plated iron pipe, don't inhale the dust.

If there is a coating that is tough, it is galva !
Fact is, it is totally unsuitable but for synthetic rollers and be ready for a bumpy ride.
Note that steel rollers will end up smoothing it. Can take some time.

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

Something like this is where i'd like to start, but a £1000 might be cutting it fine, hopefully 2k would be a capable machine.
[www.youtube.com]

[www.youtube.com]

[www.youtube.com]

[www.youtube.com]

Now, these are serious machines !

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)

So you're prefer a lathe over a mill? I don't think that fancy, slanted bed, auto tool changer lathe could ever make it to anyone's doorstep for 1K or even 2K. Can't say much about the other ones but I'd wager they'd be over 1K if they were ever sold as a commercial product.

I'm aiming for something well below 1K, I know I personally couldn't justify spending 1000 pounds on anything. I'm looking to make a machine that's under 300 pounds. So linear rails like these machines are using are a no go, so are the ball screws I think I saw in a few of them.

EDIT: not saying they're bad machines. The first one in particular looks amazing, especially for a scratch build. I'm just trying to say that's out of the price bracket I'm going for, and most of the replies that mention cost have said to try and bring it down to the cost of a 3D printer (I'm guessing the cheap variety?).

Edited 1 time(s). Last edit at 04/18/2018 05:45PM by Trakyan.

I've been looking around for a decent headstock (solution) like a sherline size without the cost, the completed 4th axis setups you can buy, i'm not sure if they could be used as the spindle, only as a rotary table, was also thinking of the RIno for an indexing table but now there are a lot of right angle worm gearboxes coming out of China, I've seen some decent large pilliow bearings 20mm id 2x for 15 quid...but I think maybe a better cheaper solution would be to get a bike bottom bracket and work to those dimensions(& possibly cheaper bits.

headstock & tailstock, X/Y with 1605 ballscrew & nice thick looking aluminium construction, add a trunion(matching the XY construction), they sell them in a sets, which cant be much worse than a sherline or taig solution.

Quote
Trakyan
So you're prefer a lathe over a mill?...

Never said that.

Just that this is the right way to build a decent CNC able to machine more than cardboard, plastic sheet, wood.
Here, as I said earlier, concrete, granite slab could be used as a base for the linear rails. Even then, you will have to spend.

Now, a CNC router as cheap as a cheap 3D printer, will be worse than the said printer.

BTW How about a summer job in a machine job to learn and earn ?

---

"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)