New CNC build with belt driven Z

[n0rth3rnlight]

I work as a salesrep for a big company selling thermically cut platesteel IE laser, plasma and gas and now my brother and i will be building a CNC from scratch. I have a few questions regarding my design choices though And hopefully you guys can help me prior to orderering the bulk of the material.

This is the Z assembly with integrated v bearings for the y axis. The Y axis will in turn be belt driven as will the Z axis. Also i know that i have to add some spacers in between the carriage and groove bearings (or thread some holes).

Now for the questions to the pros:
1: the groove bearings will ride on 2 pieces of steel and 1 set will be adjustable up and down (as have been done many times before) but do you guys think that a laser cut rail will be holding tolerances ok?
2: The Z carriage will be 1 piece and the fit will be adjusted with the adjustable v groove bearings (as excentric spacers are very expensive). What about the tolerances on the z carriage riding surface? Will it be ok you think?

Similar Threads:

• Video Belt Driven Z Axis
• New Machine Build- belt or gear driven?
• direct driven or belt driven
• belt driven CNC
• Belt driven screw ?

[ger21]

Is this for a router?
How big is the machine?

Do you plan on grinding a bevel on the steel "rails"?

[n0rth3rnlight]

Is this for a router?
How big is the machine?

Do you plan on grinding a bevel on the steel "rails"?

thanks for the reply. The machine will measure somewhere close to a 1200mmx600mm cutting area the i do not plan on grinding the rails. Is that stupid? Also yes it´s for a colt style router supposed to cut wood, plastic and small pieces of alu.

[ger21]

It's not ideal.
V-Rollers do not make the most rigid Z axis, and not grinding the edges won't help. They'll where rather quickly, and constantly need adjusting. And they probably won't wear evenly, so they'll be loose in some places and too tight in others.

I wouldn't recommend a belt drive either, but if that's what you want, try to use a gear reduction so that one revolution of the stepper motor moves the machine about 15-25mm.

[n0rth3rnlight]

It's not ideal.
V-Rollers do not make the most rigid Z axis, and not grinding the edges won't help. They'll where rather quickly, and constantly need adjusting. And they probably won't wear evenly, so they'll be loose in some places and too tight in others.

I wouldn't recommend a belt drive either, but if that's what you want, try to use a gear reduction so that one revolution of the stepper motor moves the machine about 15-25mm.

Ok thanks The Z is no problem, i can change the movement to something more ridgid than a v groove bearing, but what about for the x and y. Do you think a ungrinded laser cut rail will do for those 2 axis?
I´ve seen some people using normal angle iron and that seems to work ok no? the laser cut rail would be much more accurate than normal steel angle atleast?

[ger21]

A lot of things will work.
The answer really depends on what your expectations are. I can't tell you if it will be good enough for you.
But I'd really recommend either grinding the bevel on the rails, or purchasing the pre-ground rails for those V rollers.

[n0rth3rnlight]

You have luckily changed my mind. I asked my friends at a mechanical workshop how much it would be to brake similar rails as The ones used by platform cnc and they told me they would do it for free

Will post updated design in a few days

[wizard]

Ok thanks The Z is no problem, i can change the movement to something more ridgid than a v groove bearing, but what about for the x and y. Do you think a ungrinded laser cut rail will do for those 2 axis?

I'm going to come out and say No here, the problem being the type of usage you mentioned earlier.

I´ve seen some people using normal angle iron and that seems to work ok no? the laser cut rail would be much more accurate than normal steel angle atleast?

Why would you think that a laser cut rail would be more accurate? The potential is there of course but normally I associated laser work with light gauge metal, much of your accuracy would come from how that rail is supported.

[wizard]

thanks for the reply. The machine will measure somewhere close to a 1200mmx600mm cutting area the i do not plan on grinding the rails. Is that stupid? Also yes it´s for a colt style router supposed to cut wood, plastic and small pieces of alu.

The images you posted first ( of the Z axis) indicate a very very thin structure for a machining application. It would probably work for a plasma or gas torch but not much else. In a nut shell the plate looks thin no matter what sort of bearing you would use.

The second issue that comes to mind is that this is a large machine for a Colt style router. You would be better off with a bit more power I would think for a machine this large. Combine that with a desire to do plastics and aluminum, I'm pretty sure your machine won't meet your quality desires. However that is me, it might help to define what your expectations are here. A machine designed to cut up plywood for boat building has different quality requirements than say a machine built to machine guitar bodies.

When you first describe a machine with V rollers and belt drive I was thinking of a machine with rather course precision requirements (boat building). When you indicated a desire to machine plastics and aluminum that leaves me with the impression that you expect far more or will need more precision than you realize. If so V rollers probably aren't for you.

So as others have already stated it is really a question of you expectations. If we don't know what your expectations are then the general response will be to suggest a more robust machine. Personally I really think that you will want a far more robust machine than that indicated by the first set of drawings. In the end V rollers probably aren't the linear motion solution you want for this machine.

Remember it isn't the size of the work here that is important. Rather it is the accuracy you want to achieve and the quality of the finished work that counts. A machine that vibrates, deflects and otherwise has rigidity problems can make an even small piece of aluminum look like it was attacked by a beaver.

Finally I'm going to make another suggestion. 1200mm is almost 48" just shy of the width of a sheet of plywood here in the USA. If you go slightly bigger in the one direction say 1300mm you will cover even oversized sheet goods from the USA. The other benefit is that you can use the additional travel to machine fixturing and clamping solutions. This is something to consider as it is a marginal growth in machine size but allows for flexibility in machine usage.

[blacksmith2]

It's not ideal.
V-Rollers do not make the most rigid Z axis

I am looking Solsylva as a design starting point and the bearing implementations seem to be the weak point in the design.

ger21 What is superior to V-rollers and still in the DIY category?

[ger21]

What is superior to V-rollers and still in the DIY category?

How are you defining "DIY"?

If you want a decent machine, there's really no comparison to profile linear rails. Not only are they far stronger and much more rigid, but they're also the easiest way to build a Z axis.
You can pick up some inexpensive used ones on Ebay, or buy brand new Hiwins from BST Automation on Aliexpress at a very good price.
You might pay $150 more than using V rollers, but you'll get a far better machine.

[blacksmith2]

How are you defining "DIY"?

By DIY I am implying that I could purchase a complete machine or a kit ready to assemble but, in this case I am choosing to do most of the construction myself. Obviously, doing things like making my own ball bearings is would be foolhardy as they are relatively affordable.

As you are a long term participant in this forum, I respect the opinions of people like you and believe you have a perspective I would work years to attain. If you have seen a mechanism that you believe to be superior to V-Bearings and could be assembled with ingenuity & resourcefulness I would enjoy learning from your insight.

[jannos]

Hi.
For my last build I bought Hiwin linear rails and ballscrews from Mekanex. This is a Swedish company, and they have pretty good prices.

[UA_Iron]

You didn't say what the stroke on your Z-axis needs to be, but this isn't a difficult problem to solve.

I've wondered why people don't use linear profile rails, instead opting for making a ton of different interfacing parts and adjustments in multiple locations to solve this problem...

Automation Materials.com | MiSUMi | METAL > Plates

Spec out a 1018 annealed steel plate - Even a .393"x6"x12" steel plate blanchard ground on two sides is about $57. If you opt for the ground on two sides and precision milled on the other 4 sides, you're looking at $87.

Find your linear guide rails - ebay is a great used market, but it can be hit and miss. These do have a life span afterall.

I was able to pick up a set of brand new NSK LH25 440mm rails, two blocks, matched set with inspection sheet showing them true to each other within 2nm for around $70. I got lucky, but really any will do. Some of the more rigid type are the THK HSR and IKO LRX. You will not need a nominal linear guide size greater than 15mm. These are orders of magnitude more rigid than the V-groove type, and much more repeatable. I'm sure for your application that the HIWIN will be just fine though
Hiwin HGR 15 20 25 30mm 1M Linear Carriage Guide Rail or Slide Rail Block Select | eBay

User your linear guide as a hole template on your ground steel plate - use gauge blocks to reference rail from the side/edge of your plate - use another standoff to reference the other linear rail off of your first linear rail. Clamp guides down hard, use a punch to dimple where your threaded holes will be located.

Ball screw
1 Anti Backlash ballscrew RM1605 225mm ​C7 with ballnut End Machined for BK BF12 | eBay

support blocks:
Fixed Side BK12 and Floated Side BF12 ballscrew End Supports Ind | eBay
interfacing flange:
1 Ball Nut Housings Nut Bracket Fit SFU1604 1605 1610 RM1605 1610 Flange Nut | eBay

And a right angle adapter for your stepper motor
And a flex coupling to couple stepper to ballscrew


Use another ground plate (or just find mic6 aluminum) to interface your screw/interfacing flange to your linear guides and of course to mount your spindle on.

I'm not sure if that illustrated the process well enough, but that is it in a nutshell.

[Devastator]

I am looking Solsylva as a design starting point and the bearing implementations seem to be the weak point in the design.

V bearings are essentially the same skate bearing with a V in the outer race. They may actually be worse than the stock design of tensioned opposing bearings because they'll get lateral movement especially when they wear and you can't use preload to remove it.

You might look into these: Extended Linear Carriage with ABEC 7 Bearings - CRP102-00 | CNCRouterParts

That's better than V bearings IMO and a good middle ground between skate bearings on pipe and linear profile rails.

[wizard]

Just to add an alternative to profile rails you could use round rail and linear bearings. Supported round rails aren't bad for wood working equipment as long as you don't have high expectations for the machine. For a machine detained to machine aluminum though profile rails would be a better choice.

However All things being considered you need to understand what your expectations are. If the primary use of the machine is to rough out wood for other processes then you might do fine with cheap V bearings if you expected nothing else out of the machine. If the goal is to machine wood to near net finished product then You need to consider a more robust machine. Note I said machine, as nice as profile bearings are, to really benefit you need a suitably rigid machine.

A point here it sounds like the machines at you place of work are mostly involved in cutting with some sort of torch. Be it plasma, laser or whatever do realize that these machines don't need to be extremely rigid because there a very little in the way of reaction forces from the tools. In fact a heavy & rigid machine actually works against the machine design if rapid operation is desired. When you talk about actually cutting materials with a router or milling type spindle things change dramatically, rigidity becomes very important as does accuracy and often surface finish.

So when you mention the materials you want to machine and suggesting aluminum is in the mix you really need to define what your expectations are with that aluminum. Any serious machining of aluminum requires that you build a significantly more robust machine than one suitable for just carving up wood. This doesn't mean you can accomplish a bit of casual machining of aluminum on a machine designed as a wood router, just that you might be disappointed if you have high expectations for a low end router.

[blacksmith2]

Extended Linear Carriage with ABEC 7 Bearings - CRP102-00 | CNCRouterParts[/url]

That's better than V bearings IMO and a good middle ground between skate bearings on pipe and linear profile rails.

Interesting. What got me thinking about V Bearings as an option was the V's on CNCRouterParts pricier "Pro" series. I agree with you, CRP102-00 appears superior.

I like the flat bar idea for the simplicity. Grinding a V on two sides of 1/4 inch stock would be a challenge to get it parallel. I would have to try of course. Ha.

Thanks