Purchasing the right linear rails for a home made cnc router table

[ger21]

I guess he didn't like the V bearing system on his Evolution machine ?

My understanding was that CNC Router Parts gave him the machine for free advertising.

I saw your post earlier on Joe's forum, where you thought V bearings might be better than Hiwins?

Linear bearings are a huge improvement over V bearings. There's a reason why CNC Router Parts went from V Bearings to Linear Rails, and now Joe is doing the same with his machine.

[MajorTendonitis]

My understanding was that CNC Router Parts gave him the machine for free advertising.

I saw your post earlier on Joe's forum, where you thought V bearings might be better than Hiwins?

Linear bearings are a huge improvement over V bearings. There's a reason why CNC Router Parts went from V Bearings to Linear Rails, and now Joe is doing the same with his machine.

Yes , my thoughts were that using V bearings may be better in a dusty environment , but I'm certainly not a fan of using V bearing with angle iron .
I do like when I seen a modified Evo using V-rail instead though .

I was very tempted to go the V bearing route , but am kinda steering towards linear rail and blocks ,as I believe it will make for an easier build . When I read there's actual grease on the linear guides it sort of had me worried , but I'm getting over it .
Have been waiting patiently to see what Joe comes up with , and yes I seen his rendering

[ger21]

When I read there's actual grease on the linear guides it sort of had me worried

The grease is in the blocks, not on the rails. Although you will see some on the rails after you grease the blocks (which you need to do regularly).
My day job is running big routers, that use linear bearings. In 20 years of heavy duty use, I've never seen a bearing failure, or had a dust problem.

[MajorTendonitis]

The grease is in the blocks, not on the rails. Although you will see some on the rails after you grease the blocks (which you need to do regularly).
My day job is running big routers, that use linear bearings. In 20 years of heavy duty use, I've never seen a bearing failure, or had a dust problem.

That's good to know . Btw , you've got a cool job

[louieatienza]

Yes , my thoughts were that using V bearings may be better in a dusty environment , but I'm certainly not a fan of using V bearing with angle iron .
I do like when I seen a modified Evo using V-rail instead though .

I was very tempted to go the V bearing route , but am kinda steering towards linear rail and blocks ,as I believe it will make for an easier build . When I read there's actual grease on the linear guides it sort of had me worried , but I'm getting over it .
Have been waiting patiently to see what Joe comes up with , and yes I seen his rendering

They may be fine in a duaty environment ut their load capacity is miniscule compared to profile rail/block

[wizard]

Welcome to the site.

Guys , I was debating to build a Joe's CNC , and while Joes V bearing Evolution is all that I would probably ever require, the more I study , I'm certain I could build a cnc table myself using square block linear rails , and CNCRP pro series rack and pinion .
Sorry if my terminology isn't great yet .

What you can get away with depends upon what you expect out of the machine. V-Bearings can be a good solution for many machines and even for rooters with specific use cases. I wouldn't suggest them at all for a router to be used for general purpose work.

My design was to have a cutting area of 4' x 5' , as a 4/8 will not fit well in my garage . I'm interested in 3D carvings and signs , and fairly good repeatability. Will be using Nema34 for steppers , aluminum extrusion most likely from Vancouver , as I'm located in eastern BC Canada .

You will most certainly want something better than V-Bearings in my estimation. Aluminum extrusions are the expensive path to a machine but are a good solution if you have a limited shop. This mainly due to the ability to order your extrusions cut to size, even drill and tapped (or other machining). Personally I prefer steel.

Was hoping to get get opinions on these linear guides and blocks in this link below .


I'm not sure if this is there lower model ? I have heard where 25mm are overkill and 20mm are fine , but I would prefer to go with these , as the cost difference isn't a lot.
Also heard mention that you should exceed your cutting area by approx a foot on each axis for the linear guide length in order to compensate for the area lost to the blocks .

25 mm is most certainly overkill, even 20 mm might be considered over kill.

Your linear rails need to die sized to take into account the size of the bearings, the spread of the bearings and other obstacles that you need to clear. The size of your tooling comes into play and you may need extra space to machine any fixturing you may need to make. Plus it is nice to have an area on the Y axis to park the spindle for tool changes.

Thanks in advance all , as I've gotta start somewhere

A machine build is a lot of work. You mentioned being spaced constrained, you need to realize that a 4x5 foot machine (work area) will take up more area than the 4x5 size implies Your machine will be at least a foot longer and probably a foot wider when all is said and done. Also a moving gantry solution needs clearance on the sides and vertically. How much clearance is need vertically depends to some extent on how you expect to do dust collection. I'd do several layouts of your shop space to make sure 4x5 is really doable.

[MajorTendonitis]

Thank you again guys for the responses , and thank you Wizard for that comprehensive post , much appreciated all .
There's an amazing amount of information here , and I'm glad to be here and try to learn ,as CNC's have always intrigued me .

I will stick to linear rails , guess I'll go with 20mm . And yes , I was expecting the footprint to be at least a foot more than the cutting area .
I have a shop , 22'x26' with 10' ceilings , but it's getting busy in there .
I looked at the measurements of the cncrp pro version , as it gave me an idea of the footprint to expect .
The 4x8 was a foot to long for me to be comfortable with . A 4x6 would be fine though , so I thought build one anywhere from a 4x4 to a 4x6 cutting area.
Some days I feel as though I should just order there pro 4x4 version and be done with it , as I don't think there's going to be a lot of savings here .
But I thought it would be more fulfilling to scratch build one , would be a great learning experience, and it would be easier to correct future issues if you have a better understanding of how things work . At least I'm hoping .

And another nice thing about building one yourself , you can buy parts and start collecting them for the build , and it won't be such a financial hit all at once . As long as you don't have to do it twice that is

[minichopper6hp]

Another advantage to profile rails... The bearing block spacing doesn't have to be as far apart as other systems to get the rigidity. More travel out of the rails.

Sent from my Z981 using Tapatalk

[hanermo]

Bigger is better.
Load capacity is not the issue, rigidity is.

Lathes, for example, are typically loaded at 2% max, of their strength.
Similar to VMCs.
See the Dan Gilabert videos for good examples.

Example.
HAAS VMCs use 45 mm blocks/rails ... with a 1500 kgf thrust.
About 2% of load factor.

Max weight on table is about 4000 kg, with about 70-80 tons+ capacity on the blocks.
About 2% of load factor.

A 12" lathe with 70 mm D bearings, has about 3000 kg of load rating.
Typical max workpieces 100 kg or less.
30:1 ratio, or 3%.

I am now using 35 mm blocks on my VMC, and spaced double blocks / double rails on the front Z axis.
Rated strength 50 metric tons+, but I plan for less than 300 kgf thrust (although the 32 mm screw and servo deliver 1000+ kgf).
The extra stiffness means I can make smaller minimum moves (submicron).

Look at screw ratings.
They have a nr for stiffness=rigidity, a 32 mm screw - 3204 - K value is 54 kgf / um.
The 25 mm screw is about 25% less.

The K value means that at 20 kgf force, the screw will yield less than 1 micron, probably about 0.5 um or so.
And at 54 kgf force, it yields 1 micron.

The x, y, axis needs about 10 kgf to start moving, and maybe 5-7 kgf to keep moving.
The Y axis is about 500 kg in mass.

For a router, You wont *need* more rigidity, but it will make the axis vibrate less, and the toolbits will wear less.
Strength is not the issue.

Kaiser for example has a multi-million business making more precise and rigid toolholders.
Because going from typical 5 microns to 1-2 microns from Big Kaiser, the carbide toolbit life goes up 500%, according to them.
Similar claims and results from RegoFix, Switzerland, double-contact toolholders, Lyndex-Nikken, Techniks, Rego-fix, Kaiser etc.


Rigidity is everything.
There is a Reason I am going to bigger rails/blocks, and a vastly more rigid machine body.

Otoh, in all honesty, about 20-25 mm is all a typical home router needs.
The frames are usually bendy, like spaghetti.
Note that a 1000 kgf Bridgeport is spaghetti. 2 fingers (8 kgf) will move it 0.02-0.05 mm, easily.

For a comparison, when I stand on my VMC spindle, == 0.02 mm, with 80 kgf force, 10x less.

Because the frame section-sizes and materials are about 100x more rigid than a bridgeport.
The frame could easily support a 100 metric ton vertical load, evenly loaded, or 50 metric tons, point load, at the center.
And yet, my max peak loads will be less than 300 kgf, mostly 50 kgf or less.
About 1000:1 ratio.
Fwiw..
Hth.

[louieatienza]

Bigger is better.
Load capacity is not the issue, rigidity is.

Lathes, for example, are typically loaded at 2% max, of their strength.
Similar to VMCs.
See the Dan Gilabert videos for good examples.

Example.
HAAS VMCs use 45 mm blocks/rails ... with a 1500 kgf thrust.
About 2% of load factor.

Max weight on table is about 4000 kg, with about 70-80 tons+ capacity on the blocks.
About 2% of load factor.

A 12" lathe with 70 mm D bearings, has about 3000 kg of load rating.
Typical max workpieces 100 kg or less.
30:1 ratio, or 3%.

I am now using 35 mm blocks on my VMC, and spaced double blocks /rails on the front Z axis.
Rated strength 50 metric tons+, but I plan for less than 300 kgf thrust (although the 32 mm screw and servo deliver 1000+ kgf).

You do realize this is a router build, mainly for processing sheet goods? That same HAAS has a 30-40HP spindle and is made of cast iron. I think the Biesse routers have only 30-35mm rails on one side, cantilevering a half-ton or more gantry. Plus those machines are using servos with 15-20 times the power needed for this build. Bigger LM rails need bigger frames to bolt on to, bigger motors, bigger screws... now you've gone way out of the scope of the machine's intended use.

[MajorTendonitis]

Bigger is better.
Load capacity is not the issue, rigidity is.

The extra stiffness means I can make smaller minimum moves (submicron)

For a router, You wont *need* more rigidity, but it will make the axis vibrate less, and the toolbits will wear less.
de toolbit life goes up 500%, according to them.
Similar claims and results from RegoFix, Switzerland, double-contact toolholders, Lyndex-Nikken, Techniks, Rego-fix, Kaiser etc.


Rigidity is everything.
There is a Reason I am going to bigger rails/blocks, and a vastly more rigid machine body.

Excellent analogy

I'm the type pf person that would prefer to overdesign if I can afford to , and the price difference is very minor upgrading from 20 to 25mm .

Maybe I should rethink the size of the extruded aluminium I wanted to use also .
I was going to use 2040 for the sides of the Y axis , and use 3060 for the gantry.
Now I'm wondering if it wouldn't hurt to use 3060 for the sides of the Y axis to.
i was going to use 2040 for the middle struts

[ger21]

You're tool life won't increase by 5x, because you're machine won't be powerful enough to get anywhere near the chip loads you need to be able to see that.

As I've said what seems like 100x, now, 20mm IS overdesigning by a huge margin.

Linear rails are only as rigid as what they are bolted to. But depending on the design of the frame, you may or may not see any difference.

[louieatienza]

Excellent analogy

I'm the type pf person that would prefer to overdesign if I can afford to , and the price difference is very minor upgrading from 20 to 25mm .

Maybe I should rethink the size of the extruded aluminium I wanted to use also .
I was going to use 2040 for the sides of the Y axis , and use 3060 for the gantry.
Now I'm wondering if it wouldn't hurt to use 3060 for the sides of the Y axis to.
i was going to use 2040 for the middle struts

You'll only realize the quoted load ratings of the LM components if you mount them according to the manufacturer's specs. Otherwise you'll have to derate them based on mounting method and material. As stiff as the LM rail may feel, they will surprisingly easily telegraph whatever irregularities in the mounting. In other words, they're only as good as what you bolt them on to. The LM rail shouldn't be the main source of structural rigidity in the machine. That said I think you should consider 4590 for your main frame members.

Which I suppose brings up a point in favor of 25mm profile rail; if you do use 45 series extrusion the t slot width is 10mm, which gives a comfortable 7.5mm for the LM rail to sit on each side of the slot. With 20mm it's only 5mm.

[wizard]

Excellent analogy

I'm the type pf person that would prefer to overdesign if I can afford to , and the price difference is very minor upgrading from 20 to 25mm .

Maybe I should rethink the size of the extruded aluminium I wanted to use also .
I was going to use 2040 for the sides of the Y axis , and use 3060 for the gantry.
Now I'm wondering if it wouldn't hurt to use 3060 for the sides of the Y axis to.
i was going to use 2040 for the middle struts

At some point it will be a total waste of money to put overly large linear rails on aluminum T-Slot extrusions. Larger linear rails wound be that much stiffer if the substrate they are mounted on can't properly support the rails. I'm not sure about your axis naming as I generally call the gantry "Y", but the one big advantage you have on the X is that you can throw in plenty of support and reinforcement for the X axis frame. That support can even be added after the fact if things are too flimsy. I'd certainly go larger that 2040, don't forget you need room on the beam for the rails and drive components which is another factors suggesting larger extrusions. Also how that frame is supported is a factor also.

It is the Gantry where you really need to be careful as that beam has to function properly when supported at only the two ends. Here is where you really want to invest the money to do the beam right. For a 4+ foot span I'd recommend a steel beam even if you build most of the machine in aluminum. I'd also suggest reading the stickies where one of the threads dives deeply into beam design for a gantry.

[wizard]

Linear rails are only as rigid as what they are bolted to. But depending on the design of the frame, you may or may not see any difference.

This pretty much completely sums it up! Further lightweight aluminum extrusions (almost all of the commonly available T-Slotted extrusions), are less than optimal for supporting linear rails once you focus on rigidity.

[MajorTendonitis]

This pretty much completely sums it up! Further lightweight aluminum extrusions (almost all of the commonly available T-Slotted extrusions), are less than optimal for supporting linear rails once you focus on rigidity.

Well my concern is , working with a steel gantry would most likely be a pita for me . I'm not trying to take the easy way out , just trying to be realistic.
I have big drill press and can make a jig to make sure all my holes are lines up properly, but I'm not sure about the rest , like would the holes be strong enough after I tapped them out etc .
After I look at Probotix's cnc design , I'm sure a piece of 3060 is going to be 10 times more rigid then there system .

[MajorTendonitis]

I'm not sure about your axis naming as I generally call the gantry "Y", but the one big advantage you have on the X is that you can throw in plenty of support .

Sorry as my terminology isn't this great yet . When I refer to the X axis , I mean the main beam that's holding the spindle . I refer to the sides as the Y axis


A concern I have is trying to get the centre of gravity right for the spindle . I like how cncrp has the horizontal 3060 of the gantry secured on top of the vertical sides .
In my case I was going to secure the 3060 to the front of the vertical extrusion, but secure the vertical extrusion towards the back of the plate , which will help a bit .
I need to fire up my laptop and use sketchup to illustrate better .

[ger21]

Just because companies sell CNC machines, doesn't necessarily mean they know the best way to build them. One reason that people build their own machines, is because they can build better machines than they can buy, in a lot of cases. CNC routers are not complicated.
A lot of CNC Routerparts design decisions are done for ease of assembly, and low cost.

I like how cncrp has the horizontal 3060 of the gantry...............

They use 80160, which is a little bigger than 3060.

[MajorTendonitis]

Just because companies sell CNC machines, doesn't necessarily mean they know the best way to build them. One reason that people build their own machines, is because they can build better machines than they can buy, in a lot of cases. CNC routers are not complicated.
A lot of CNC Routerparts design decisions are done for ease of assembly, and low cost.




They use 80160, which is a little bigger than 3060.

Thanks Gerald , as I wasn't aware of that .
Well that could be an issue , as that's not available where I'm shopping . I figured if I built my own , I was hoping to make a machine that was more solid than the pro version from cncrp .
That's part of reason I wanted to go with 25mm guide rails and larger 3060 for the sides where the guide rail will be attached to the Y carragies.
Guess this may be more challenging than I first thought

[louieatienza]

Sorry as my terminology isn't this great yet . When I refer to the X axis , I mean the main beam that's holding the spindle . I refer to the sides as the Y axis


A concern I have is trying to get the centre of gravity right for the spindle . I like how cncrp has the horizontal 3060 of the gantry secured on top of the vertical sides .
In my case I was going to secure the 3060 to the front of the vertical extrusion, but secure the vertical extrusion towards the back of the plate , which will help a bit .
I need to fire up my laptop and use sketchup to illustrate better .

Naming the axes depends on how you set it up in your controller. Usually the longest axis is called X, with the shorter axis Y and the vertical axis Z. Sometimes however, the gantry is the longer axis and the part is loaded facing the gantry, and the machine is set up such that the gantry is the X axis. It doesn't matter one bit. If you really needed to, you can make a profile for each orientation. It's a lot easier, however to just rotate the part in CAD or CAM!

As long as you have the spindle center somewhere between the footprint of the gantry bearing blocks, I think you'll be fine. You may find that getting the CG over the center of the bearing blocks moves the spindle past that bearing footprint, which would increase the amount of rail needed to get the travel you want.

Looking at the CNCRP PRO kits, it looks that Gerry's spot on about that gantry beam being 80160... If you held a piece of 3060 in your hand it would become patently clear that it won't do for a 5' span.

One thing to consider when designing your machine would be to incorporate raised rails on the long axis; that is, the extrusion, linear guides and motion components are above the surface of the table. This would eliminate having to engineer the uprights, which should lead to a more stable and rigid gantry. Reach-in to the machine may prove a little more challenging, but the overall machine height could be lowered slightly to accommodate that.