Desktop CNC advice request

Robin2:
This a Forum for helping people with their own Arduino projects. It is not a Linux forum.

I suspect you will get a lot more help on some of the many CNC forums.

Are you aware that there is a program called LinuxCNC ?

...r

This is why I asked in this part of the forum. After all, it's description is

Feel free to talk about anything and everything in this board.

The many CNC forums you speak of are typically geared toward professionals ( or at least seem to be ) with high end machines. I've not, to this point, found one that has a section for hobbyists/noobs that really gets any traffic. For instance, the one I found this morning. The last post/reply was back in May of this year.

As for LinuxCNC, I have seen mention if it in my research. However, most, if not all, I've read on it were sales pitches, for lack of a better term.

I will indeed seek more information on that particular program. Thanks!

ballscrewbob:
If you are going to buy a small machine off the shelf so to speak please make sure there are NO 3d printed parts involved. You may have to pay a little extra for such a machine but you will thank me later LOL.

Once you get above a certain size the G-CODE control mechanisms change from mostly GRBL based to MACH based.

However once you have played with a smaller machine moving up to a MACH based machine is quite a short learning curve as you will already have most of the basics down pat.

Better also to avoid those that have the Arduino NANO based control board approach and go for the UNO based.

Currently use my stuff mostly on hardwood so I know if I slow them down even just a small amount they would also be fine for Aluminium.

Don't know what your budget is but a decent size laser of 5W or greater is a fantastic addition and opens up the machine to a lot of new areas.

Don't use Linux here so cannot say much in that aspect.

Bob

This pic might get your juices flowing LOL

Thank you very much!

I had read elsewhere to avoid machines with 3D printed parts. But nothing I have read has explained why. I can assume that it has to do with the parts being too flexible or break/wear out.

I had no clue there was a need to use Mach for larger machines. I figured that the computing would be pretty much the same regardless if size. But there will time to figure that out later.

And the laser is something I have thought about. And I may go that route to start with in my learning process. Just thinking with a laser there are no expensive cutters/bits to break!

Did you make that clock? Awesome!

Robin2:
Are you aware that there is a program called LinuxCNC ?

Just went to the LinuxCNC website. Looks like I have some reading to do. Thanks again for bringing it up. Not sure why I had not went to the source on that one until now.

Dang...
I knew you were going to ask a couple of "WHY ?"

OK the 3d parts from China and most others will fail for a few reasons.
Infill not sufficient
Cheapest plastic used for the extruder or even re-used plastic used to make the plastic.
30-60% less compression strength.
There are more but those are the main reasons to avoid 3d parts.

As for GRBl to MACH it is pretty simple and more to do with power requirements and motor drivers.
You would be taking that big leap from hobby to semi pro and looking for longer term durability.
That said I probably have a few thousand hours on my hobby machine and the same on the two laser based home brew machines (printers and plywood and chewing gum approach).

Lasers can bring thier own minor issues in that FLAT OBJECTS are mostly the the things you can do until you get into 4th axis or fancy surface tracking. (can be a steep learning curve)

Yes The clock is only one of many wonderful objects I made that keep the wife gasping and allowing me to spend so much time with Arduino and my machines. It was simply a combination of open source artwork / clipart put together in a couple of hours and then machined into a chunk of red maple.

Bob.

ballscrewbob:
However once you have played with a smaller machine moving up to a MACH based machine is quite a short learning curve

Just out of curiosity did you use GRBL or MACH3 to make your clock face?

It seems to cost money and I can't find out how you interface it to hardware - presumably it is just a tad more modern than LinuxCNC and is not expecting you to have cobweb infested PC with a parallel port :slight_smile:

...R

ballscrewbob:
Yes The clock is only one of many wonderful objects I made that keep the wife gasping and allowing me to spend so much time with Arduino and my machines. It was simply a combination of open source artwork / clipart put together in a couple of hours and then machined into a chunk of red maple.

I still think it's awesome. More evidence to support the "Happy Wife = Happy Life" rule!

PS.
Thanks for the explanations!!

Hi R2.

Was going to ask how you are doing with your set up.

I have not yet made the move to MACH based machines but am pretty close as I need a larger machine.
But currently all done using GRBL based machines.

All the simple UNO with a CNC shield on top.
Gone through about 5 driver boards but almost all of my own stupidity by hogging down the machines.
But they are pennies anyway really.

I have a copy of ARTCAM (now defunct) that makes most of the work easy to do simply based on artwork.
There are a few open source packages but I have taken to this pretty well and it came with the first machine I got off the shelf so not looking a gift horse in the mouth.

Most of the newer MACH is USB based or you can even step it up to standalone too.
We all find different packages more to our liking for ease of learning.
I know you like your linux and I like my windows.

Bob.

Side note for Danger is that with cheap diamond bits from China you can also lightly engrave on stone ! Or with a 10/15W laser also write directly on quite a few types of stone.

ballscrewbob:
Side note for Danger is that with cheap diamond bits from China you can also lightly engrave on stone ! Or with a 10/15W laser also write directly on quite a few types of stone.

Hadn't even given working with stone a thought. That's pretty cool though. I'll keep that in mind. Thanks for sharing.

@Danger.

WARNING...These things can get very addictive and almost any object within arms reach can become a target for these little machines...."I wonder IF.....Oh I know what would be pretty cool to do...Hey just stick your hand on here for a few minutes....Oh I saw some wooden skids just down the road i will be right back...Hmmm now how can I modify this little $#!% to make it work better.

Dang it ballscrewbob!

I'm over 50. Bladder control isn't a problem yet. But ya ain't gotta test it like that! That was friggin funny.

ballscrewbob:
Was going to ask how you are doing with your set up.

As usual I got sidetracked. I have spent several very enjoyable weeks building a Python program that is a GUI front end for the OpenSCAD 3D modelling program. Unless you use OpenSCAD I doubt if it will have any interest but there is a short video here.

...R

Oh and did I mention making paint stencils with the laser.

Well, in my search for a decent machine, I am really leaning towards the MillRight M3 minus the Router and mount. I really don't want to use a wood router to do the job of a spindle. Not to mention the noise a wood router makes. I know spindles can be loud as well.

Anyone heard anything bad abought the M3? Or with MillRight in general? Not real keen on the frame being made of MDF. But that could be upgraded later on to something more durable and stiffer.

Not keen on the belt drive either to be honest or the Vee rails.
(explanation) Real CNC machines tend to always use direct drive or a small pulley to direct drive not a full belt for the whole travel. It is also less noise overall too. Also there is no SLACK to take up that can introduce some positional errors.

If that's about the size you want to start with then I would say look for something that has direct drive from the stepper and runs on at least 12mm linear rails. (most about that size will)

I know the chinese ones look crappy but the frames tend to be aluminium and very sturdy.
The "TEE" bolt frame can take many additions and lend themselves well to modifications (speaking from experience LOL)

Mine has had some plywood added and an extra PSU for the better spindle and some extra controls are mounted on that along with fans (old computer) on each side in PUSH PULL mode to get the fumes and fine dust to one side for easier clean up...

I agree on no router approach for the smaller machines as the spindles are quieter.
But you should look for a minimum 500W (300W would be an ok starter) with an ER11 collet on it.
And ER11 chucks up to 1/4 and 6 MM although you would find a lot of the smaller bits use 3.175 mm regardless of size. You WILL break some of the smaller tools until you get to grips with speeds and feeds so having extras is a bonus.

Don't expect to use STD router tools on small stuff unless you are quite careful.

Mine is a bit of a Frankenmill with the modifications but it does what I ask of it and I can swap it back to its original config without too much hassle apart from the Y and Z axis that failed (3d printed).

Bob.

Addendum.. In another life I installed ballscrews to CNC lathes and mills and was also a service tech for a couple of CNC machine suppliers hence the nic.

ballscrewbob:
Addendum.. In another life I installed ballscrews to CNC lathes and mills and was also a service tech for a couple of CNC machine suppliers hence the nic.

That makes me appreciate your insight all the more!

I looked at the Chinese CNC machines like this one. And was putting thought into how to recreate those printed parts in a stronger material. Or how to swap them with better parts.

Aside from that, and the cheap spindle, those machines seem somewhat solid. There's even a vid on YT where a fella machines aluminum with his. No way near as fast as the MillRight M3 was shown to do though. That said, the MillRight had a top of the line spindle and a high quality mill/bit. (Still working on terminology)

The ability to machine aluminum is not mandatory for me. Just a bonus.

The machine I linked above was not the one I was eyeballing. It has plastic uprights. This is similar or maybe even the same as what I was looking at.

Those grey plastic lumps are what will fail eventually.
Luckily for me I saw it coming and used the machine to make its own replacement parts somewhat.
A few suppliers offer upgrades to those parts and for the Z its often worth it as you will probably end up with more Z travel.
That extra Z travel can come in quite useful too.

If you have somebody who can machine HDPE or aluminium that would make great replacement parts.
If not and you are handy (and accurate) with wood then a suitable hardwood makes just as good a replacement.
My spare parts came from a hardwood skid and I used the machine to do most of the work but simplified that "U" shaped bit to just a block that carries the bearings and lead screw. Then added two right angle plates I found at the hardware store to carry the rest of the assembly.

But see all those empty T slots on the side that just beg for something to be mounted like a nice piece of ply so you can move the power supply and electronics to a better area and give yourself room to add some extra controls and gadgets.
Also acts as a guard sort of to keep a lot of the sawdust more inside the machine.

Mine was a 3025 with a 50mm Z now its a 3025 with 170 Z and a plethora of electronic bolt ons.
Once you have one in your hands then you will quickly see which bits could have been made better or what you need to add on.

BTW if you can be patient enough with waiting times for delivery then look at Aliexpress or even Ebay as Amazon tends to be more expensive for this game all round. Even my tooling comes from Aliexpress.

For tooling try avoid dual flutes and go with 4 flute preferably without the hole in the end for coolant and if you can find them the ones with slightly offset cutting edges on the bottom.
I got some 4mm ones a while back and am amazed at how much deeper I can go for a pass and how much quicker too. Not broken one of those yet either...Rough out with end mills and finish off with a ball end.
Although on occasion I finish with 1.0 mm END mills but you have to be careful with those.

BTW despite my CNC background it was all mechanical and I never programmed them so all this was a learning curve for me.

Will throw a couple of pics up when I get a moment that may explain better than I do.

Bob.

I was actually contemplating using epoxy resin to cast blocks large enough to machine down. Stuff sets hard as a rock and cuts and drills pretty well. Would have to test if a 3018 can mill it though. I guess a lot would depend on the spindle and tooling chosen.

I linked to the one on Amazon only because their links tend to stay active longer. eBay was where I was looking though. And you're correct. Much cheaper.

And I would love to see how you modded the Z axis. The Z travel is IMO the one of the biggest cons to these machines besides the plastic bits. Of course, my opinion is based on very, very limited knowledge.

Some poor pics of the mods.

The black brackets found at a hardware store. The lower one was modified to allow the spindle body to pass by

Extra parts used.

2x optical rod 8mm x 200mm long
2x flange mount for 8mm rod
2x end mount for 8mm rod
1x 8mm leadscrew 250 mm long
1x 8mm flange bearing.

Rest was scrounged from the Z failure.
You can also see some rubber strip I used to help damp the spindle noise but that has to be positioned carefully otherwise you will throw out the the tooling alignment.
Just had enough room on the machine to get most of the depth needed for the bearing holes etc in the wood with a long series 6mm end mill, and finished off on a regular drill stand.

Other pic shows the plywood dust catching and fan.
It will take two fans if I want as there is a plate over the second hole.

I would be wary of EPOXY RESIN as it has a tendency to fracture under stress. You will encounter stress in more than one way LOL.

Bob

ballscrewbob:
I would be wary of EPOXY RESIN as it has a tendency to fracture under stress. You will encounter stress in more than one way LOL.

Bob

I really should have thought of that. I built an electric guitar for my son last Christmas and finished it with clear epoxy. Makes for one heck of a mirror shine when all buffed out. But he bumped it pretty hard going through a doorway and it popped a big chunk out. Was a pretty easy repair. Way easier than a paint finish.

I'll probably start out similar to yours by using a hard wood. Once I figure everything out I should say.

Just did the "make offer" thing on a 3018 on eBay. Have to wait and see if they'll go for it. Their original price was well below anything on Amazon. And I asked for another 30% off. We'll see.