Any CNC hobbyist here?
 

I'm a complete noob and I'm starting to play with stepper motors, drivers and Arduino ... have a couple of basic questions.

Mini-CNC-Mill project slowly progressing.

Y-axis done. Was a pain re-machining a $200 Chinese x-y table to have close enough tolerances to be useful. Motor drives are up and working, Arduino controller working and G-code software working. Still a ton to do.

https://i.imgur.com/Iq4hwn8.jpg

What are you using for a stepper driver?

I had to put the brakes on my project a few years back when it just got too big for the space I had. It’s a 5’x2’x1’ machine made from IAI Linear Actuators I somehow acquired out of a Toyota plant. Geckodrive stepper driver (undoubtedly outdated now) and I forget what size/torque my steppers are. Sad thing is it’s already setup with servos and I have their driver, but their software is proprietary so I have to hack the servos out.

Please keep us updated or do a build log post to keep some of us motivated!

I was looking at doing one of these set ups but my buddy whose a CNC guy keeps talking me
Out of it. Tells me it’s a waste of time and energy. So I’ll be watching hoping you prove him wrong .

Entropy there are a ton of Chinesium driver boxes for sale now, just size them to the application, these are DM542T .. not sure if that's a part number or has something to do with what driver chip it's using. Super simple, hook up, about $25 each. Most of my drivers and steppers come from Amazon via SteppersOnline.

Turbo.. yeah it's really hit and miss as far as being any real world use. Rigidity and mass of course is a problem, Z axis and spindle can be pain. I'll have maybe $1000 invested when it's all done and I'd be extremely happy if it will handle a 1/4" end mill on aluminum.
Mine diverges from the standard DIY build in that I'm not using those cheesy aluminum linear bearing slides ... this one has real cast iron dovetails and gibs.

I expect it to "function" just fine, I'll be quite surprised if it actually does a good job and leaves a good finish. If it's actually useful I'll consider that an unexpected bonus ... it's primarily a learning process for me. I have lots of experience in CAD modeling, and some experience in the electronics ... but am a beginner when it comes to CAM, G-Code etc. So ... yeah, a bit of a science project for me.

But I would certainly agree ... if a person is expecting to build a production quality, profitable or even really useful DIY CNC mill .... that's a very tall order and a person would be better off biting the bullet and buying a used CNC machine.

Not really a hobby, been machining for 35+ years.
I always thought it would be cool to have a mini mill to do scale model stuff.
Never thought about building one though.
If you are looking into any type of linear bearing check out
Hiwin. We use them all the time in the stuff we build and they are pretty accurate.

Heres some shots of the latest. Think 18ft caliper.
X & Y on this one. Have done X, Y & Z on others and the next one we are adding a head so they can do angle measurements. Gets laser calibrated to +/- .002.
The scale it uses is imbedded in the linear rails and I think there is a way to make it talk back to a controler for location.
These are 20 & 25mm, but I think they have 10mm and possibly 5mm.

Pav ... very nice ... I've already been thinking about the next design and I'm pretty much coming to the conclusion that bolt on linear bearings would be the way to go. Would allow me to choose the mass/bass I want without having to worry about machining ways in a massive chunk of cast iron. I'm limited to what my BP will handle ... but could still throw a 200 lb chunk of cast iron on the table and true it up for bolt on bearing surfaces. Would also make fabing a Z axis way easier.

Yep, the intended use for this would be things like scale model work, maybe useful small parts in aluminum and brass ... I'd be super happy if I can get results accurate to .002" in the x-y.

Just about have the x-axis ball screw ready to mount up. This shows the underside of the table that I had to machine out to clear the ball-nut and adapter. Also had to re-cut the dovetails as they were out .005" in the z direction (in China that's a tenth :)) Got them within +/- .0005 on the clapped out old BP.

https://i.imgur.com/83ouZrj.jpg

You got me thinking now. Just happen to have some ball screws and may dream up a table top mill for tinkering.
Hard to fathom the amout of parts I have done on a Bridgeport. Way to easy nowadays to walk right by ours and jump right into one of our CNC's.
Even for simple parts.

You've got me thinking about those linear rails for a Z axis .. so those things are designed to take machining loads, not just positioning like for a plasma/laser etc?

Seems like bolting some to a machined chunk of grey cast iron would be pretty simple.

They are built to varying designs. Some to accept crazy high loads and accuracy.
I will snag a book at work and look more into it.
Otherwise just look for Hiwin brand.
I can probably scource some. May even have some laying around depending in size needed.

I went to their website, they have a very complete listing of products and specs. I actually see some of their smaller stuff for sale on Amazon. Might have to pick one up to eyeball it. Appreciate the tip ... has to be way better than trying to machine and maintain cast iron dovetail ways.

Update: Well jeez, after researching these linear bearings I would have gone in a completely different direction with my build. Evidently yes they are commonly used in high end machine tools and are capable of handling whatever loads I would be likely to generate. A bit pricier but worth it for the versatility and ease of use. Huge selection available, I notice McMaster has butt loads of them. Will definitely be my goto for the Z axis .. I could fab and square up my own column and mount the bearings to it.

PAV .. thank you very much, and ... I hate you :)

Well, my current build will at least be very useful for learning purposes.

What length rail do you need John? I have various linear rails (all THK I think) that are just taking up space. They might be longer than what you need for a Z but who knows.

I had a long 5’ one I kept tripping over so I mounted it to the wall and fabricated a mount for my TIG welder and all accessories. When I need to weld I drop it down to chest level, when I’m not welding I send it up to the ceiling lol.

If I have a shorter one that works for you then you’re definitely welcome to it.

A trick we use is when mounting the rail or linear blocks, make a shallow pocket for the linear blocks and a shallow key for the rails for a nice tight fit.
Something around .100 deep depending on chamfers that are on them.
It helps keep shock loads from moving the block and rail and eases the lining up of them when assembling.

Entrophy .. at this point I still have to decide what width rails I want to use ... probably small, like 15mm wide. Price increases dramatically with size when it comes to the blocks. Length would be in the area of 8" (200mm), two of them. I'll definitely let you know when I settle on a size. Thanks!

Pav ... that thought did occur to me .. figured I could machine both pockets on the BP at the same time to ensure they are parallel (I'll be using two rails, four blocks) then installation would be a breeze ... put the rail in place and use a transfer punch to mark the bolt holes.

Please forgive my random reaction but seems vibration control would be a big consideration for you and it hasn't been mentioned. I can imagine the behavior of a ball bearing slide would be a lot different than machined cast iron slides (stiffness, mass, and damping). Do any of your CAD tools have vibration analysis modules?

I am a little sensitive to this after dealing with a few brinelling and resonance issues in my past life.

In the past vibration control for machine tools was basically handled by massive grey iron castings ... obviously not a solution for a desktop machine. These days I think vibration is less of a consideration because of the very high spindle speeds .. 10k - 50k rpm. Resonance at those rpms is probably at a frequency above what would cause problems with accuracy or surface finish. Compared to spindle rpms in 2k-3k range common in the 50's. I could be wrong because it's certainly not my field of expertise, but I think the higher the frequencies the smaller the mass required to dampen them.

But ... for this project, at present I am just trying to include as much mass as I reasonably can and have some level of portability. And trying to pay as much attention as I can to geometry to lessen the forces that would create vibration. I've got stress, flow and electrical simulation ... no vibration analysis. I did once write a small program for Helmholtz resonance calculations in a fluid system ... but I don't think it's applicable :)

I'm very surprised by the capabilities of these linear bearings, I guess they have totally replaced cast iron ways in CNC applications. Even the machines with iron ways are all using Turcite surfaces now. Whole new world compared to my 1957 Bridgeport.

That makes some sense. I was definitely thinking drilling and milling and massive old-school machine tools!

10K to 50K spindle speed sounds like pure terror. I'd need a big green light as I wouldn't hear it running!

Here is a link to a video of the first testing. Just the X - Y axis. I built a little stylis to hold a plotting pen than I can raise and lower manually to simulated the Z axis.

Drawn image is just part of a logo I did for another project that I shrunk down, exported as a DXF file to Fusion 360 which generated the G-code, which I sent to the Arduino controller with Universal G-code Sender.

I can retrace the drawing multiple times with absolutely zero deviation between the plots. I think I'm within my target accuracy of .001" ... but don't expect that to hold up long under use with the dovetail ways. New design using linear guides already in the design phase :)

https://youtu.be/TSn9qUQzucA