Upgrade Guidance

Hello everyone,

I am in most interested in cutting aluminum at the moment but am also trying to get into some acrylic sign making.

I am aware of the various upgrades that people recommend for milling aluminum, my question is more on guidance of what upgrade is my money better spent on at this time.

Machine info:
Shapeoko 2 ugraded to Xcarve(no x-controller, no single piece x-axis)
arduino uno controller
nema 17 motors(54 oz-in I think)
60 min x-axis stiffening mod
about to have aluminum T-track bed
Dewalt 611

My thoughts so far:
Motors: I think that the motors are probably one of the weakest areas of my machine currently. I cut 3/4" aluminum and all is smooth until I get deeper into the material, then I start to see a lot of shaking of the whole unit and can hear the increased stress. I know I need to look into better methods of cutting instead of just cutting a trench around the profile of my parts, but I am figuring that out in fusion currently.

Upgrade options:
179 oz-in nema 23’s(would this require new 9mm belts?)
Linear Rail Z-axis
Ball screws for axis instead of belts
X-controller(I don’t know of any real benefit of this over the arduino other than a cleaner setup)

Any thoughts on which upgrade my money should go to first is appreciated. I think I am leaning towards the motors currently.

Thank you,


FWIW - this is my experience from a diy CNC based on supported rails/screw drive.
Some info about it here: My DIY-carve

1 - Get Nema23’s in the 269oz range for all axis where the motor case will fit. Machine some Nema17-23 adapters if required.

2 - Get new drivers so you can take the motors to their optimal current limit (or get an Xcontroller)

These two mods are the two most signifigant changes to my diy CNC. (Started with Nema17’s and A4988/DRV8825 drivers)

Stronger belts are preferred as the increased motor torque may make the belts your next weak link. Screw drive will also be a big factor.

Video of mine in action (diy):

Fusion360 is very powerful once you get your head around it, it did take me a little while… :smiley:

Thanks for the reply. There is a lot of good info in the link you posted. I appreciate it.

I certainly can’t pull the trigger on all of these upgrades at once. But it looks like the consensus so far is that the motors would be the first place to put money into, which is what I was thinking anyways.

I wasn’t thinking that I would need a new controller for the motors though.

looking for the TB6600 on amazon results in items that are much less than $100. I must not be seeing the right thing. or do I also need 4 drivers to go with the controller?

I have a 24v 500 watt psu(https://www.amazon.com/gp/product/B01IU8QBCO/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1)

I think that this is adequate for the stronger steppers. is this accurate?

Thanks for the input.

I am not looking to be able to mill aluminum in a matter of minutes. I am more concerned with being able to start it and have enough confidence in the machine to walk away and come back later to a successful cut and not have to worry and babysit it for an hour.

This - you have the Arduino and PSU so all you need is 4 new drivers as Phil suggest, without the added power the Nema23´s won’t be reaching their potensial.
I went same way as Phil did, just slightly different drivers, and the difference was night and day. No understatement… :slight_smile:
4x TB6600 required

1 Like

where can I get 4 of these motors for $80? on amazon, they are ~$140 for 4 of them. on stepperonline.com they are $70 +$55 shipping.


that was the problem, I didn’t notice the country option. of course, they have no stock for the US.

Also check the shaft diameter of the Nema 23 you choose, Nema17´s are usually less.

With 4 new motors and 4 new drivers, you have about $200 invested which will be a massive upgrade, power wise.

looking at the print of the nema 17 on inventables site, it looks like it does have a smaller shaft, therefore I will need new pulleys as well.

If your goal is cut quality improvement instead of cut speed improvement does that chance the answer?

The load is higher so a stronger system is less prone to loose steps. However, to successfully carve metals another major factor is rigidity. The more rigid the structure is the better.

Another factor at play is how you carve. Easel is lacking in that regard as more powerful CAM-suites can generate specific kinds of tool paths, well adapted to the task at hand.
With Easel I can’t efficiently move aluminium, so I need to make it spend its time, fast and shallow.
With Fusion360 (or similar CAM´s) I can make different strategies better suited.
I have done 9mm depth of cut with a 3mm bit (1/8"-ish) at 750mm/min (30ipm). The cut strategy limited the step over to a certain extent in order to not overload my system.

Here is an image of a stress test I performed a while back, 8 test pockets with 1mm increment depths running full depth at 600mm/min and 0.8mm step over. Carve finished in 12min with Fusion360, CAM 2D Adaptive clearing:


What length is that belt in the link you provided? I cant tell on the page unless im just blind. is it by the foot?

Ok so I have decided on doing the motor upgrade with the new drivers, pulleys and 9mm belts while I’m at it.

The more I use my machine and the motors miss steps the more irritated I get with it.

I just need the steppers to be in stock for USA shipment.

Is there a bigger belt for the z-axis? I haven’t seen a 9mm closed belt yet.

1 Like

I had nema 17s on my machine at first and jumped to 180 ish in/oz motors on my arduino and gshield. I do plan to build a new controller at some point but the new motors were a dramatic improvement even with the reduced amperage available with the gshield. When I go new controller it should be even better.

1 Like

It will, my experience is that the new drivers (for me) was by far he largest increase in performance/resistance to loosing steps. I dont think I have lost a single step since that upgrade, knock on wood!
Larger motor with small drivers will have a decent holding torque (at idle) but dont reach their potensial until they get full power.
I have a Nema23 running my X-axis (rated at 269oz @2,8A) limited to 2,6A and have tried to hold the axis back by hand/force and I failed to make it stall. (Voltage 24V and running half step)
I am using ACME rods. FWIW :slight_smile:

Everything has been ordered except a belt for the z axis. Need to find one if it exists.

Excited for the improvement!!!

1 Like

SO I haven’t found a GT3 closed belt for the z-axis, so I am considering going to direct drive.

Where do I find the coupler nut to connect the z-axis acme rod to the motor?

I understand that I also need to make an adapter plate to mount the motor to, does anyone have a plate already that I could share the file with me?

Your best bet for the Z is to grab a linear direct drive unit from cnc4newbies. They are super rigid and come with everything you need. If you message them and ask they will drill the mounting holes for the x carriage and the Dewalt spindle mount for you at no cost. Only thing you will have to do is decide on a way to mount your z homing switch. I drilled and tapped mine for M3 screws but some people use epoxy with decent results.


1 Like

That’s a 3rd gen GT belt with 2mm pitch and 80 teeth (like the Inventables one), 9mm width.

That’s a 3rd gen GT belt with 3mm ptich and about the same length, 9mm width.