Motor selection for kit build

[KentW491]

Looking to select Nema23 motors for this C-beam kit: https://www.makerstore.com.au/product/kit-otbv1-s-1015/

It seems my options are:
* Standard steppers (e.g. https://www.makerstore.com.au/produc...nema23-635-ht/)
* Closed loop steppers (e.g. https://www.makerstore.com.au/produc...cl-2nm-hss-al/)
* Servos with integrated controller (e.g. https://www.makerstore.com.au/product/elec-ihsv57-180/)

For starters I was thinking of using the AXBB-E motion controller with UCCNC (or Mach4) to get up and running relatively quickly, with the possibility of moving to LinuxCNC with a Mesa board in the future if I get around to ATC, 4th axis, etc.

I have more or less ruled out standard steppers as the engineer in me wants closed loop control.

My understanding is that from the motion controller perspective, the motors still appear as open loop with step and direction control, however the alarm output from the stepper driver / servo controller could be used to stop the machine in the case position is lost.

The price difference between closed loop stepper and integrated servo doesn't appear to be much, so would there be any reason for me to selected the closed loop stepper over an integrated servo? The only possible advantage I can see would be the holding torque of the stepper.

Should I be considering integrated servo for all three axes, or potentially using closed loop stepper on the Z axis for the holding torque?


Extra question for the future - if I were to move to LinuxCNC, given that the control loop of the integrated servo is hidden from LinuxCNC, am I going to be missing out on anything of note?

[Al_The_Man]

Incidentally, NEMA rating is a physical mounting size, Not a torque definition, e.g. NEMA23 motors can range from 42 to 214 oz-in torque rating.

[ardenum]

servo's are generally rated at lower torque values then steppers but perform much better in the real world.

[ger21]

servo's are generally rated at lower torque values then steppers but perform much better in the real world.

Yes. Servos will have much lower rated torque, but will have more torque at much higher speeds. You'd like need to use gear reduction on the servo, to fully utilize it's performance capabilities.

On a small, simple machine like that, I'd just get simple open loop steppers. They'll give you the same performance as closed loop steppers, and are simpler, and cheaper.

For starters I was thinking of using the AXBB-E motion controller with UCCNC (or Mach4) to get up and running relatively quickly, with the possibility of moving to LinuxCNC with a Mesa board in the future if I get around to ATC, 4th axis, etc.

UCCNC can control ATC's, rotary axis, and probably anything else you'd want to do. If you may have a need for more I/O down the road, spend a little more on a UC300ETH up front, and you may find you'll never want to switch to LinuxCNC.

I look at it like this. Say you have an open loop machine that loses steps. What you do, is configure the machine to run within the motors capabilities.
Now say you have a closed loop machine that loses position and faults the drives. What you do, is configure the machine to run within the motors capabilities.
If both machine have identical spec motors, then they'll both run reliably at a given speed, and neither will run reliably above a certain speed.

[peteeng]

Hi Kent - I think using servos on that machine is over capitalising the machine. The hybrid steppers will give you a bit more speed then the std steppers. Given the construction of the machine I'd go std steppers get through the learning curve then use it to build a better machine from the ground up vs trying to develop the Outback further. UCCNC and a UC300 will give you a lot of flexibility and as Gerry says you probably won't look back. Peter

[KentW491]

Thanks all for the feedback, particularly on the suggestion to not over spec the motors for the machine. The intention of this build was to start with something basic to learn on yet still functional, and then a new more advanced build in the future. Just need to avoid the spec creep.

I'll definitely check out the UC300. Glad to hear that UCCNC will cover everything.

Just one last question on the motors - would the servos with integrated controller in fact be a worse choice for the Z axis over either stepper options due to the holding torque? Note that the structure is driven by leadscrew, with the spindle weight somewhere in the 5-9kg range (will open a new thread on that can of worms). I could envisage the Z axis needing to constantly correct with the servo as it fights the weight of the spindle whereas the steppers could hold the spindle solidly in place.

Not having built a mill before I'm unsure if this is an accurate guess or a non-issue in reality.

[rodw]

the machine is designed for steppers. So use steppers. You will have enough to learn without having to stuff around with tuning servos. Servos on a leadscrew driven machine would be a waste.
Linuxcnc will offer more flexibility long term but might be a longer learning curve. I chose this path and would not do it any other way.
Yes with linuxcnc it is possible to add linear scales and use an inner and outer position loop for greater accuracy.
But as said, drive the motors within their limits and open loop steppers will be fine.

[peteeng]

Hi Kent - What do you want to cut with this machine. You mention "Mill" it is not a mill and won't do well cutting metal. Peter

[KentW491]

Hi Peter,

Thanks for your input. Yes, sorry, I meant router rather than mill - an artefact of limited experience in this area.

The short answer to your question is aluminium; however taking a step back a bit to provide some background and context:

Our main product requires a few aluminium pieces in low volume. These are currently produced externally.
I personally have an interest in learning and getting a CNC up and running, and combined with the ability to manufacture the parts in house I can justify putting a little money towards it, ~$5000 AUD.

The parts include:
* Heatsink to transport heat from some electronics to an external enclosure, starts with 20x20mm bar and needs machining on two surfaces with a few steps to clear certain components and match the enclosure profile, and two taped holes
* Mounting bracket of similar size / complexity to the heat sink above
* End plates from ~1.6mm aluminium sheet that fit to an aluminium extrusion with cut outs for various connectors. Whilst the majority of the endplates will still be made externally, from time to time we need to customise the product which means different cut outs for the connectors. Would be great to be able to produce these on the machine rather than our current solution of 3d printing them.

With limits on space, time, and budget; a CNC kit seemed to be a good balance between budget and designing everything from scratch and buying something off the shelf. I was targeting the ~1.5x1m size as we don't really have space for anything bigger, and that sort of size opens up other possibilities for working with other softer materials such as creating jigs, etc.

The kit previously linked on Makerstore comes to ~$4k AUD when paired back to standard steppers and 1.5kW spindle.
The two issues I see from this (and my other) thread so far are:
* This kit may not be able to meet the requirements mentioned above
* Spending a similar amount elsewhere may give me a better performing machine

My thought was to start with something basic, then potentially move to a different, more advanced in the future after gaining the experience if the need / desire / budget arises.

It seems that working with aluminium is towards the upper limits of a router. My guess is that within my price range at least, machining the surfaces heatsink and cutting out end plates will be OK, but machining an M3 thread or similar may be too much of an ask. Creating the thread is a 'nice to have, not required' feature.

I had selected the Makerstore kit as the C beam they use had a denser profile than other similar options I came across. I did ask for anyone local to me (Sydney) who may be interested in letting me see / talk about the machine, but they weren't able to provide any contact details.

I'd appreciate any further input, potentially saving me a costly mistake.

[KentW491]

I've been doing a lot more reading of this forum and research since my last post and can see that several places offer similar designs with the lighter c-beam section (relative to the makerstore c beam heavy), but with ball screw instead of lead screw for much lower prices. I could even get the buy the base kit and replace the c beam with the c beam heavy and still end up ahead.

The question remains, is this type of machine with ball screw really capable of working with aluminium?
Would the c beam heavy from makerstore add any significant stiffness to the machine?


Also from looking at the other posts, with a lot more time and experience, I could turn out a much more capable machine. Something to look forward to in the future...

[peteeng]

Hi Kent - turn your private messages on. Peter

[KentW491]

I wish I knew how.
There seems to be some oddities with the settings on this platform.

Selecting "account settings" from the profile icon on the top left takes me to a screen that says I have zero forum posts across the header (obviously not true). Clicking through the various account settings doesn't seem to give me an option for controlling PMs. Forum settings doesn't appear to have any relevant option, and if I set my time zone correctly it seems to always revert to GMT after saving the settings.

I've sent a message to you via the Messenger section (icon to the left of profile picture at the top), and also attempted to add you as a contact by hovering over you name on a forum post.

[ardenum]

it will cut aluminum but slowly and loudly.

[peteeng]

Hi Kent - if you look at this thread #589 you will find me. Peter

https://www.cnczone.com/forums/diy-c...ml#post2504340

[KentW491]

Gotcha.

Also sorted out the forum PM mess. In case anyone has a similar issue - whilst the posts are available at both of the following URLs, the account settings pages are different! PM control can be done from the first, I was accessing from the latter.
https://www.cnczone.com/forums
https://en.industryarena.com/forum

[joeavaerage]

Hi,
I rather suspect that the machine is insufficiently rigid for regular and cost effective manufacture of aluminum parts. Such a machine would be great for wood and plastics,
engraving of all types, and the very slow painstaking cutting of aluminum or brass.

The rigidity required to cut aluminum is about five times that is required for wood.

I would be less concerned with the steppers/closed loop steppers/servo debate and do some serious research on the suitability of the machine frame.

Craig

[dazp1976]

I'll definitely check out the UC300. Glad to hear that UCCNC will cover everything.

The UC300eth is a very good controller and the ucbb breakout board is great.
However, in your case. The AXBB-E will be an easier route as a starter as it's all in one. It is based on the 300eth chip. If any more IO are needed you can add on to port #3.

For the serious user, the UC300eth-MAX from Cncroom is the one to get (at a cost!).