Building a DIY CNC

Woodshop
#1

Figured woodshop was the best place for this, but, mods, feel free to move if it fits better elsewhere :slight_smile:

So, I got the green light from wife to build my own CNC, and I’m hoping others here have gone before me. The goal is at least 1’x2’ workspace (would like 2’x4’), and I’m bouncing between something like Bob’s CNC or Ooznest, though I’ll likely be sourcing components myself to save cash.

So, the questions. Drawer slides, linear rails, skate bearings? I’m leaning toward linear rail kits as they’re affordable and seem to yield good results. Chain drive or lead screw? Lead screw for z-axis seems best, but read a bunch of different opinions on X/Y options. I have some NEMA17 steppers from Tanners and an arduino cnc driver board. Should I make the jump to NEMA23? I’ll mostly be milling MDF and softer woods. For the spindle, I’m looking at using a Dewalt DW660 like Bob CNC uses.

I have about $500 to spend on this project.

#2

Never built one myself but I’ve seen a lot of builds use https://www.cncrouterparts.com/.

1 Like
#3

Linear rails for sure, belt or gear drive for X/Y. Leadscrews have whip issues when you try and run them too quickly or make them too long. NEMA17s are wimpy, go for the 23s.

1 Like
#4

I am about to pull the trigger on a Sienci CNC kit ($400 + router), but it only has a work area about the size of a sheet of paper. Would be very interested in someone putting together a build kit list and directions…I’d also help out if there was any way to do so. Any chance of turning this into a class?

2 Likes
#5

@Chris_Wischkowsky built his. He may be able to shed light on do’s & do nots.

2 Likes
#6

Thanks! Didn’t know that about lead screws. Belt seems the easiest to access after digging into gear drives :slight_smile: I’m curious on your recommendations for the motors. I’ve seen a bunch of machines/kits with 17s, but I know there’s a lot of variation in motors.

#7

nice! I hadn’t seen that one. My timing sucks for taking on a project like this as our first kid is due Jan 16th :slight_smile: I’m tempted to do a kit like that one to simplify the entry point :slight_smile:

Digging around, Sienci gives you a full build list in keeping with their open source mantra: https://staging3.sienci.com/resources/

1 Like
#8

Whip is reduced significantly if the screw is tied down on the end.

#9

All the lead screws I’m looking at have mounts on both ends… From what I understand, you can get higher feed rates with belts, but the simplicity of 1 stepper per axis driving a lead screw is very appealing.

#10

Screws give you better precision than belts at much lower feedrates.

#11

My buddy recently made one using the plans and parts from https://www.vicious1.com/ . It works awesome.

1 Like
#12

A philosophy:
more bigger is more better
Get bigger steppers

#13

@brsims Dang it! Now I have to put the stops on my Sienci purchase and explore this!!!

2 Likes
#14

Yeah…I’m now considering the MPCNC…dang it. I was all ready to buy the Sienci.

Anyone want to coordinate a group build? It’s bigger than the Sienci, but when built to recommended specs can fit in a trunk or small SUV. I’d be willing to put in a couple of weekends with folks.

#15

I’m torn between MPCNC and going fully DIY. The more I dig into this, the more I like bits and pieces of different kits. I’d love a more robust router, but I love the idea of something like this as a first build!

#16

Definitely like the bigger, ergo better approach… my wallet is less enthused by that approach :slight_smile:

#17

Come now, 23s aren’t that much more expensive.

#18

Save Your Ears

The type of motors you choose depend on what your want to do with it. Since you are limited to a $500 build I would highly recommend you spec out a small machine and fit as much quality as you can onto it. Nema23’s are not that expensive but your drivers will get larger and so will your power supply. Big motors are not a replacement for good mechanics. I would imagine you could properly spec out a build using nema 17’s.

Rigidity is key, the frame is the most important part, that is where you should be spending the majority of your budget and where you don’t have budget, your time.

The worst linear guides are generally better than the best linear rails. The only reason to not use linear guides is if you simply don’t have the budget. Linear rails are ok. Do not use wheels or plain ball bearings.

Spend money on good connectors, cable, and wire. This cost will be a lot higher than you think, but if you cheap out… you will regret it.

4 Likes
#19

True, but with a smaller budget, I’m having to figure out the best allocation of resources. Not to worry, my hope is that this machine is a gateway drug :smile:

#20

Thanks for jumping into the fray! Oh so many questions now!

First on the spindle, I like the idea of a spindle instead of a router, especially since learning that a fair amount of people have issues with the collets on the trim routers. Any concerns on the limited RPMs on the one you linked to? I thought you’d want to be able to take it up to 20k RPM for some bits.

Agreed on the frame/mechanics, and from what I’ve seen/read, NEMA17s at least 1.8a should have enough beef for a home cnc.

I hadn’t even looked into linear guides! Are they ok to be turned on their side? Do you recommend these for the Z axis as well? Nearly every Z axis I’ve seen has been a linear rail like this.

Lastly since @Clayton and I are both looking at the MPCNC… thoughts? Not really looking for something to compete with the Multicam, but not wanting to cheap out too badly either.