Starting to design out my build, looking for critique

[Chrono]

Hello everybody!

I have been reading posts here and designing my own CNC for a while now. I have been sitting on this and not continued for quite some time, but I now have the opportunity to access a lathe and a well-equipped workshop for a couple of months, so I want to finally move this forward.

Before I go into the design, I want to quickly state what I want to achieve with this machine: I want a relatively large (read: Around 1m x 0.7m x 0.1m) area machine to cut wood, plastics and aluminium. At the same time I really want to constrain the cost as much as possible, not wanting to go over a cost of 4000€ (at this point I should probably mention that I am located in Germany).

Now with the information I have gathered, this is the design i came up with to satisfy these requirements:





The Profile is Motedis 45x90S B-Type Slot Aluminium Profile, The rails are HIWIN rails and I am planning on using ballscrews for the motion. For the Rails and Screws I plan on buying a bundle on Aliexpress which will likely dictate the exact dimensions of the machine.

Now to the questions I am still having:

• How would you go about connecting together the Profiles? Would angles suffice, or do you use something else there?
• What would be the best way to mount the Rail Bearings? The green elements in my Model are just placeholders, since I struggle to find a good way to mount them.
• How would I place Drag Chains on this design, to protect the cables? Would I need to make sure to have extra flat space to route these?
• I was planning on using NEMA 34s with Timing belts to drive the screws. Would I need two of them for the Y Axis?
• The Spindle Plate hanging on the Gantry looks huge to me, compared to the rest of the machine. Did I misdesign this?
• Is a machine like this even capable of milling Aluminium? Would it be able to cut thin (not more than 10mm) steel sheets?
• And finally, is this design even any good? I mostly worked off the information I found in various posts on this forum, especially for the gantry design, but I have no idea if the result has any glancing flaw.

Thank you for any help!

[AndyUK]

Is a machine like this even capable of milling Aluminium? Would it be able to cut thin (not more than 10mm) steel sheets?

Others will be able to say, but should be able to take light cuts in Ali. Steel is probably out of the question - for that you're better off converting an existing mill.

The Spindle Plate hanging on the Gantry looks huge to me, compared to the rest of the machine. Did I misdesign this?

Can't see it in the renders? The carriage plates look reasonable - but obviously there isn't a Z axis yet which will change things.

I was planning on using NEMA 34s with Timing belts to drive the screws. Would I need two of them for the Y Axis?

This size of machine should be fine with Nema 23, but that's just a mounting size. On your long axis, you can either go with a pair, or you can use a single motor with a timing belt connecting the screws. If a belt across the CNC isn't a massive inconvenience to you, my vote is for the belt - otherwise dual drive. There is a motor calculations spreadsheet on the forum which will help direct you to which motors to use.

How would you go about connecting together the Profiles? Would angles suffice, or do you use something else there?

You can tap the ends of the profiles and stick large bolts in, you can use angles, or you can use T-nuts. There are probably other ways - look at a Profile supplier's website for inspiration (e.g. KJN).

And finally, is this design even any good? I mostly worked off the information I found in various posts on this forum, especially for the gantry design, but I have no idea if the result has any glancing flaw.

Looks like you're on the right lines to me - I'd carry on by improving the detail of the CAD.

[Washout]

• What would be the best way to mount the Rail Bearings? The green elements in my Model are just placeholders, since I struggle to find a good way to mount them.

Hi Chrono

Andy beat me to many bits of good advice, but I'll add:

Whilst some of the details are missing the green plate method, provided the plates are thick enough should be a good approach, as you will be able to adjust the plates to get the ballscrews aligned properly. In fact being able to make adjustments for alignment of ballscrews and stepper/servo motor mounting should be your aim when designing the mounting points/plates (with rigidity of course :-) ).

Your Z Axis carriage/box will likely need some side bracing of some kind to stiffen it - could be as easy as some triangular corner plates to tie the right angled joint together.

Hope that helps.


Chris

[AndyUK]

Andy beat me to many bits of good advice

There is a reason I picked which questions to answer and which to avoid!

Washout makes a good point about the adjustment of things. I like your ballscrew placement idea on the gantry, but it doesn't leave a lot of room for height errors - the location of the top plate will be defined by the carriages, which may or may not co-inside with the height of your ballnut support. You may want to design with say, a 5mm shim under the BK/BF12, so that you can adjust the shims to get the right height. This is why the ballscrew on mine is on the back of the L shape, it allows me to adjust for height relative to the carriages - but leads to a slightly more flimsy 3-sides of a box shape.

[Chrono]

Thank you two for the replies! I have been implementing some of the things you were talking about in my spare time, but much more needs to be done yet.



What I changed:

• I added angles and connector plates. My plan is to mount the spoilboard on the angles.
• I added the Z-Axis. Currently, due to the different heights of the Rails and the Screw-sleds, the plate holding the spindle has a weird U-Shape. I could cut into the Z-Carriage and sink the screw into the gantry a bit to make this flat, but thus sacrifice gantry rigidity. Any thoughts on this?
• I also added Angles to help support the Z-Carriage, like Washout recommended.

What I am yet to do:

• I understand the thought about keeping the mounting plates adjustable, thus I will design out those green plates to make actual mounts. The same for the back bearings.
• I want to add motors to both the Gantry for the Y-Axis and the Z-Axis. For the Y Axis I thought about mounting them in the sidepanel holding up the gantry. For the Z-Axis I thought about lengthening the top plate and hanging the motor there. Would this be a problem in terms of placing extra torque on the Gantry?

I also have a couple more questions:

• Should I put in two more profiles to brace the bed in the Y direction? Crossbracing will be hard using only Aluminium profile
• How thick should the different Alu plate parts, like the bearing plates, sidepanels and Z-Carriage be? Currently, I have all those set at 20mm thickness.
• What about endstops? I have access to a 3D printer, can I use that to print mounts for the endstop sensors, or would those be too flimsy? I do not see them having to take much force, but maybe I am missing something here.

On your long axis, you can either go with a pair, or you can use a single motor with a timing belt connecting the screws. If a belt across the CNC isn't a massive inconvenience to you, my vote is for the belt - otherwise dual drive. There is a motor calculations spreadsheet on the forum which will help direct you to which motors to use.

What would be the advantage using a single motor? I am guessing the, more or less, guaranteed synchronization of the ballscrews?

I like your ballscrew placement idea on the gantry, but it doesn't leave a lot of room for height errors - the location of the top plate will be defined by the carriages, which may or may not co-inside with the height of your ballnut support. You may want to design with say, a 5mm shim under the BK/BF12, so that you can adjust the shims to get the right height. This is why the ballscrew on mine is on the back of the L shape, it allows me to adjust for height relative to the carriages - but leads to a slightly more flimsy 3-sides of a box shape.

Regarding this, would it also work if I were to flip-down the part of the Z-Carriage in the back that currently is mounted to the Screw Sled? That way I could freely adjust the height. I would only probably have to do it behind the current place where the screw is mounted, since I would hardly be able to actually screw anything in the other way around. Also, regarding the washers, would I just simply place washers between the sled and what I want to mount? The question arises, since that would massively decrease the area of contact between the two.

Again, thank you for the Replies!
-Chrono

[AndyUK]

I want to add motors to both the Gantry for the Y-Axis and the Z-Axis. For the Y Axis I thought about mounting them in the sidepanel holding up the gantry. For the Z-Axis I thought about lengthening the top plate and hanging the motor there. Would this be a problem in terms of placing extra torque on the Gantry?

As in, hanging the motor off the back of the carriage, the opposite side to the spindle? Don't think it'll be a problem in terms of torque on the gantry (else you're not going to be cutting much!), infact it may even help to bring your centre of gravity closer to the middle of the X axis bearings, which is where it should ideally be. The only issue I see is space, you won't be able to place the CNC up against something without loosing travel on X.

I added the Z-Axis. Currently, due to the different heights of the Rails and the Screw-sleds, the plate holding the spindle has a weird U-Shape. I could cut into the Z-Carriage and sink the screw into the gantry a bit to make this flat, but thus sacrifice gantry rigidity. Any thoughts on this?

I'd certainly be putting a second pair of bearings on that Z axis and lengthening the rail to retain your travel. Remember that the moment on that axis is Force x Distance to pivot - currently the pivot is the centre of your single bearing, the forces are at either end of the same bearing. With two bearings, the pivot becomes the point between them, and the forces are acting on the bottom of the bottom one, and the top of the top one. Your deflection in that direction will be significantly reduced.

How thick should the different Alu plate parts, like the bearing plates, sidepanels and Z-Carriage be? Currently, I have all those set at 20mm thickness.

What about endstops? I have access to a 3D printer, can I use that to print mounts for the endstop sensors, or would those be too flimsy? I do not see them having to take much force, but maybe I am missing something here.

20mm Ali plates for the main structures sound reasonable. Theres a spreadsheet on the forum for calculating gantry deflections that you should be looking at, and it'll give you an idea of how to approximate the cutting forces involved, and what deflection you'll expect from different materials. Regards endstops, I wouldn't be worrying too much about them just yet, but a 3D printed mount for the limit switches sounds like a good idea. With any luck they'll cut the motion before they take too much force, but having them break rather than damaging the linear motion components or your gantry sounds good ;)

What would be the advantage using a single motor? I am guessing the, more or less, guaranteed synchronization of the ballscrews?

Spot on - once you've set the relationship between the two screws with the timing belt, you don't need to worry about the gantry racking, or homing the motors independently every time you use the CNC to ensure a perpendicular X and Y. If you miss steps, they're missed on both sides - which is probably safer. Then theres the reduced complexity in limit and homing switches, and you don't have to buy yourself an extra servo/stepper driver which is £50-100 right there. You can also use a 4 axis controller and still have a spare axis for development later on.

... basically, I'm mildly regretting going for dual motors. But I didn't want the timing belt across the entire machine, so I'm paying the price!

Regarding this, would it also work if I were to flip-down the part of the Z-Carriage in the back that currently is mounted to the Screw Sled? That way I could freely adjust the height. I would only probably have to do it behind the current place where the screw is mounted, since I would hardly be able to actually screw anything in the other way around. Also, regarding the washers, would I just simply place washers between the sled and what I want to mount? The question arises, since that would massively decrease the area of contact between the two.

That sounds like how I've designed mine (see the build log).

Washers? I'd use some shim material which maintained the surface area.

[Chrono]

Sorry for the long radio silence, things have been quite busy here and I could barely find the time to work on this.

I took the advice with the Y-Axis Ballscrew and the Z-Axis sleds to heart and changed the corresponding geometry:


On the to-do-list are the following things:

• Add a mounting point for the Z-Axis NEMA 34 behind where the Y-Axis ballscrew currently connects to the Y-Axis sled.
• Add an "Omega-Drive" belt structure to the front to run the two X-Axis Ballscrews from one NEMA 34 stepper. The "Omega-Drive" structure is there to maximize surface contact between Belt and Gears
• Finally add that back ballscrew-bearing plate to connect it to the base structure
• Add a Spot for the Y-Axis Motor to sit. Due to clearence with the Y-Axis sled I am still unsure where to put that, but maybe even below the X-Axis rails on the bottom?

Do any of these plans throw up a red flag from something I have not considered yet?

I have also been talking to the mechanical engineers in my workplace, who came up with a plethora of ideas to alter the design, so I am seeking a second opinion on some of these:

• Substitute the Aluminium Profile for Steel Box section and weld it. The warping should be controllable in a machine this size, is what I am hearing.
• Change the Y-Axis gantry to a design that has a Profile up top and one below with the ballscrew sandwiched in between. The rails would go to the front and both profiles would be held together at the sides, like they are now, and with an extra plate at the back. Alternatively manufacture this out of steel section.
• And now to the weirdest one to me: Mount the gantry to the X-Axis like it is now, but only on one side. Then have the other side be out of thin steel with only a loose connection to the gantry beam, thus only inhibiting rotation, but not translation. This would "solve" the problem of the gantry being over-constrained, since the second beam would only brace against torque, but allow the gantry to move over it, in case the rails are not exactly parallel. I know that in theory this makes sense, but it still seems counter-intuitive to me in terms of rigidity.

In the mean time I will work on getting the steps on my to-do list done and report back once I am done with that!

[JAZZCNC]

Hi chrono,

Wouldn't go with ESS as the backup is rubbish and your limited to mach3/4. The uc330eth or 400 are both much better options.

I've not read all your thread but i saw you where struggling regards ballscrews and sizing them. So here's a suggestion which might help that I've built many machines using.
Use 2020 with 2:1 ratio. This gives same speed as 10mm pitch but halfs the screw speed so less chance of whipping screws. It also doubles the torque from motors.
Also very easy to increase velocity if needed by changing ratio.

Next suggestion is to buy drives that accept AC then you only need to buy transformer. That said building toroidal DC psu isn't difficult and don't let anyone tell you different but they are the best type for Cnc machine that moves quickly with high speed directional changes like routers make where drives have to handle dynamo affect of steppers.



Sent from my SM-T580 using Tapatalk

[Chrono]

Hi chrono,

Wouldn't go with ESS as the backup is rubbish and your limited to mach3/4. The uc330eth or 400 are both much better options.

Understood. I've been scouring the forum and the web, but it seems my search-fu is a bit off, since I could not really find a conclusive answer on the differences of the UC300eth, UC400eth and AXBB-E, especially since CNCDrive's web pages are kind of scarce on the matter. As far as I am able to see, the 300 just has more IO than the 400, withe the AXBB-E has the same amount of IO as the 400, but a built in Breakout in a nicer package? Judging from this post, it seems that I can use all the IO that I can get, which would tip me in favour of the UC300eth.

Also, you mentioned being locked into Mach 3/4, as a negative. What other (better?) alternatives are there, and why is that?

I've not read all your thread but i saw you where struggling regards ballscrews and sizing them. So here's a suggestion which might help that I've built many machines using.
Use 2020 with 2:1 ratio. This gives same speed as 10mm pitch but halfs the screw speed so less chance of whipping screws. It also doubles the torque from motors.
Also very easy to increase velocity if needed by changing ratio.

I guess I just ignored the existence of 2020 Ballscrews :D I calculated it through in the motor spreadsheet, and, asuming I did all the correct modifications in terms of doubled motor RPM and doubled torque, I could actually now hit 12m/s rapids with the same motor, compared to 7.5m/s before. Thank you for the recommendation!

Next suggestion is to buy drives that accept AC then you only need to buy transformer. That said building toroidal DC psu isn't difficult and don't let anyone tell you different but they are the best type for Cnc machine that moves quickly with high speed directional changes like routers make where drives have to handle dynamo affect of steppers.

Building a PSU should be fine, for the most part, but just so I can make myself a complete picture, do you have any recommendations of AC Drivers that are worth it? I am kind of missing an equivalent to what mouser or arrow are in the electronics component space, where you have one page that you can search for all components you could ever need, all in one place. I never know where to look for the parts involved.

On a different note, I have seen Jog and MPG pendants mentioned and presented on shop pages. What do those things do exactly? What I have found repeatedly on the topic, is that they are used for manual operation of the machine, however I was under the impression that this would also be possible directly from the control software. Am I mistaken here, or is there a big advantage to using such pendants?

[JAZZCNC]

Understood. I've been scouring the forum and the web, but it seems my search-fu is a bit off, since I could not really find a conclusive answer on the differences of the UC300eth, UC400eth and AXBB-E, especially since CNCDrive's web pages are kind of scarce on the matter. As far as I am able to see, the 300 just has more IO than the 400, withe the AXBB-E has the same amount of IO as the 400, but a built in Breakout in a nicer package? Judging from this post, it seems that I can use all the IO that I can get, which would tip me in favour of the UC300eth.

Also, you mentioned being locked into Mach 3/4, as a negative. What other (better?) alternatives are there, and why is that?

Yes your spot on with differences between these boards. The UC300 is the better option if you need lots of I/O. Coupled with a Good Bob that can accept 24v signals it's an excellent setup.

The alternative to Mach3/4 is UCCNC. Whether it's better or not is matter of opinion. The trajectory planner or put in simpler terms the way it creates motion is better and smoother than Mach3, can't speak for Mach4 because it's so full of bugs don't trust it and won't use it.
Now mach3 as it's advantages in that it's got massive following and is much easier to write custom scripts etc and customise to your exact needs.
What can say is both work well and will do what 99% of people want straight out the box.

I guess I just ignored the existence of 2020 Ballscrews :D I calculated it through in the motor spreadsheet, and, asuming I did all the correct modifications in terms of doubled motor RPM and doubled torque, I could actually now hit 12m/s rapids with the same motor, compared to 7.5m/s before. Thank you for the recommendation!

If by12m/s your meaning 12 meters per second then your massively off with your settings. 12 Meters per Min is more realistic figure.

To be honest that spread sheet is load of rubbish and spits out unrealistic figures. Real world is much different due to the many variables that affect machine performance.

IBuilding a PSU should be fine, for the most part, but just so I can make myself a complete picture, do you have any recommendations of AC Drivers that are worth it?

If you can find any the AM882-H are AC version. They are hard to find now thou.
If your on a budget these are not bad drives which i've fitted on several machines without any troubles and good performance.
https://www.ebay.co.uk/itm/2DM860H-2....c100148.m2813

I mostly tend to fit Closed loop hybrid steppers these days and many of them can be driven with AC or DC.

On a different note, I have seen Jog and MPG pendants mentioned and presented on shop pages. What do those things do exactly? What I have found repeatedly on the topic, is that they are used for manual operation of the machine, however I was under the impression that this would also be possible directly from the control software. Am I mistaken here, or is there a big advantage to using such pendants?

MPG / pendant is just a hand held box that lets you stand at the machine and Jog it arround rather than using the PC keyboard. On large machine this is very helpfull when setting tools or find edges of material etc as walking back to PC to make move is a pain.