[GPT]

Hi all.

Having become enthused by reading these forums, I'm gearing up to start my first CNC router project.

The plan is for a (hopefully) cheap MDF build largely inspired by elements from designs such as JGRO, Joe's 2006, Grunblau and many others.

I've attached a few sketches mostly to show what I'm planning to use for the linear motion system. It's a bit of a departure from the more common 90 degree skate bearing on angle or pipe rails but I expect it should work.

The table (1200mm by 600mm) will be a bolted together torsion box with the counter sunk holes for the outside beams milled as short slots. That ought to allow me to adjust the tension on the horizontal rollers by altering the position of the outside beams/rails.

I've shown the bearings as being on a separate block for the purpose of illustration but they will be attached directly through the gantry. The lower vertical bearings will be fixed while the upper vertical bearings will be through diagonal slots for adjusting up against the underside of the top rail.

The upper and lower rails will be steel angle while the central rail will be steel flat bar. I've shown the angle as being attached to the table top and bottom but it may make more sense to attach it to the side beams to help stiffen them.

The motion for the Y and Z axis will be similarly built.

For motors I'm looking at a kit from Xories on ebay (anyone heard of them?). 3 x 260 oz in motors, controller and power supply which all looks kind of plug and play for about $270 AUD delivered.

I'll initially use 1/2 inch BSW threaded rod for lead screws (and yes, I can see you all cringe). I'll expect I'll upgrade later, or change my mind when I get to that stage, but the cost of decent multi start acme screws looks to be about the same as the cost of the motors and electronics. I'm not expecting great speeds - 300 rpm give 25 inches / minute and I don't imagine I could reasonably expect to go much faster than that.

Did I also mention the part about being cheap?

(Perhaps when the powers that be see the machine running on the BSW screws I'll get authorization to by decent screws )

Anyway, I'd appreciate any comments and advice people might have.

Cheers

Gerard


--------------------------------------
I also plan to use this thread to keep a running tally of what I spend on the build.
--------------------------------------
Sheet of 6mm MDF $20
Sheet of 18mm MDF $33
50 1/4 x 5 bolts $20
50 bearings $72
100 cross dowels and bolts $64
--------------------------------------
Total (so far) $209
--------------------------------------

[spyderxxx]

You might want to consider Joe's plans for the hybrid. The access to the forum is worth the price. I bought the plans over a year ago and have finally started a build. It's very similar to CarveOne's here and eliminates the costly Vbearings.

Ed

[ger21]

Since when does Joe's Hybrid not use V bearings? All the ones pictured on his website do, as do all the ones I've ever seen.

[spyderxxx]

Joe's and all variations of his do. Mine doesn't

http://www.cnczone.com/forums/cnc_wo...ial_build.html

Just saying the forum is worth the cost of admission. Sorry for the confusion.

Ed

[GPT]

Hi

Joe's 2006, the JGRO and first Grunblau machine use what I call the 'offset' skate bearing design. That is sets of skate bearings bolted at 90 degrees to each other on some aluminum or steel angle and riding on tube (Joe's 2006 and JGRO) or angle (Grunblau) rails.

The Joe's 4 x 4 hybrid and the BlackToe / BlackFoot do use V bearings, but I've already brought the skate type bearings so I'll stick with that.

While the offset skate bearings obviously work (as I've seen plenty of videos of the Joe's 2006 and JGRO variants cutting out stuff on YouTube) I have always been bothered by the way I imagine some element of the force of the gantry weight acting axially through the skate bearings (which are really only designed for radial loads).

I've opted to arrange my bearings so that the that they only experience radial loads.

The rail arrangement on the Joe's 2006 also looks like it requires a degree of cutting precision I'm not sure I'm capable of. There does seem to be some adjustment to be made in the position of the bearing blocks but I don't think I'd manage to get the get the pipe carrier holes on the torsion box frame cut accurately enough.

The JGRO design does provide lots of scope for adjusting rail positions, and I considered using those plans.

In the end what I'm trying to do is build something that:

[] can mostly be constructed from simple rectangles I can cut on a table saw
[] only applies radial load to the bearings
[] provides plenty of scope to adjust out any sloppiness in my construction methods

I've been finishing up a different project but I expect to start cutting MDF some time next weekend.

I'm tackling the torsion box first and I'll post some pics when I get that done.

Cheers

Gerard

[ger21]

I have always been bothered by the way I imagine some element of the force of the gantry weight acting axially through the skate bearings (which are really only designed for radial loads).

I'm with you 100%. When trying to preload the bearings on these systems, you quickly find that the bearings stop rolling smoothly, with out much peload at all. This hinders the ability to get a really "tight" setup.
4 years ago I tried to come up with an alternative.
http://www.cnczone.com/forums/linear...ound_pipe.html

[zool]

Why are you redesigning the wheel??

There are many well designed CNC machines out there, including all of Joe's and Ahern's and CNC Tool Camp's [none of which I have a thing to do with], that have well developed parts lists, build sequences, and all the bugs worked out, as well as forums with people's experiences listed that will help you along in a manner you cannot envision. Though CNC Zone folks do not lack in knowledge and experience, the machine specific forums are more focused and you will find answers rather than having to ask so many questions.

Further, if its money you want to conserve that is fine, but you might consider that you could be wasting time, which is more valuable than money.

You might consider that by using say, a Joe or Ahern design, and spending what it takes do a high-quality build [do not cheap out .. alignment and calibration will be a horrid experience if you do], you could be making money as that will allow you to be more productive faster, or it will allow you to get to the underlying reason [a hobby] for having a tool like a CNC machine.

The time you save from using another's design can be spend on perfecting your CAD and CAM skills rather than wasting time on frustration and disappointment. CAD and CAM skills are what make a CNC machine valuable, not the machine itself.

I built one from scratch, a scaled down version of a commercial design. I was helped by a machinist, and an electrical engineer [and I paid them for their help .. gladly too!!], both of whom are WAY smarter than I and have WAY more experience than I could hope for in two lifetimes.

The biggest thing I learned is that it is better to gain access to other's talents and experience with well spent money, rather than trying to reinvent the wheel, which many end up doing badly and in a horridly time consuming and time wasteful fashion, that is frustration.

However, if you just have to flatten one or the other or both sides of your head, than have a go at it.

[Do not take my comments as being disrespectful or denigrating of your skills or abilities, these are menat to widening the discussion of your wanting to build a CNC machine. In building another's design, you will get more than enough experience ... at times you will find yourself with extra. I am currently building another's design [Shop Droid], and it ain't all beer and skittles!!]

[GPT]

Hi ger21

4 years ago I tried to come up with an alternative.
http://www.cnczone.com/forums/linear...ound_pipe.html

I had a look at your thread and like the design.

I'll have a bit of a re-think think about how I'm setting up the bearings. Perhaps welding the two steel angles of your design to a plate and attaching that to the gantry might be an idea?

Rather than your shaped rail I could probably attach some square RHS to the side of the torsion box to engage the horizontal bearings. That will also stiffen the sides of the table.

Thanks for pointing that out.

Gerard

[GPT]

Hi Zool

Why are you redesigning the wheel??

One of my hobbies is making wooden medieval scientific instruments like clocks, astrolabes, sundials, nocturnals and the like.

I do all my designs in Autocad so a CNC machine will, hopefully, enable me to more quickly turn the designs into finished pieces (this is all for fun, by the way, not profit).

As to redesigning the wheel, I'm not sure I'm doing that. The elements I've incorporated into my design have all come from links to machines I've found on these forums (so in a sense I can't really say the design is all mine anyway).

The bearing arrangement I'm planning on using can be found in adds I've seen on the forums for commercial linear carriages from CNC Router parts, and is similar to the design shown in ger21's linked thread.

So while my design in it's entirety is not proven, all the elements seem to be.

And I guess I find it more interesting to work with my own design.

It's like making wooden clocks. I spent tens of hours on the one below and ended up with a clock that, although it works and looks kind of neat, it's no where near as accurate as a $5 clock from the market. So should I have just bought an electric clock?

I'm into the idea of making a CNC machine more than the idea of having one. Sure, it will get used when it's made. But I'll probably enjoy the frustrating tinkering required to get it going.

And don't worry about offending me. I'm happy for people to say they think I might not be going about things in the best way possible

Gerard

[louieatienza]

Cool design, though the main problem I see is that if those two rails are not exactly parallel and flat, those trucks will bind in spots and be loose in spots. You'll have inconsistent cuts.

I think instead of having angled slots for preload adjustment, it might me easier to make eccentric bolts to adjust preload. It could be as easy as taking rod a drilling an offset hole and threading it.

Although you mention the problems with axial loading on bearing setups like Joe's2006, JGRO, and Solsylva, there is an inherent "springiness" to the setups that allow for tight preload, yet stil allow for very slight rail misalignment.

[zool]

GPT:

You clearly are a recidivist tinkerer based on your statement: "One of my hobbies is making wooden medieval scientific instruments like clocks, astrolabes, sundials, nocturnals and the like. "

You are in good company as CNC Zone is thick with recidivist tinkerers.

So do have fun!!!

Z.

[GPT]

Cool design, though the main problem I see is that if those two rails are not exactly parallel and flat, those trucks will bind in spots and be loose in spots. You'll have inconsistent cuts.

I think instead of having angled slots for preload adjustment, it might me easier to make eccentric bolts to adjust preload. It could be as easy as taking rod a drilling an offset hole and threading it.

Although you mention the problems with axial loading on bearing setups like Joe's2006, JGRO, and Solsylva, there is an inherent "springiness" to the setups that allow for tight preload, yet stil allow for very slight rail misalignment.

I'm hoping that cutting all the framework for the torsion box on a table saw without adjusting the fence will mean that the skins, and hence the rails, will be parallel. Having said that, I'll have to actually put it together to see how parallel I can make it

I agree that some kind of more robust preload adjustment mechanism is probably in order. You mentioned the springiness inherent in the offset skate bearing arrangement which has set me to thinking about using actual stiff springs in some fashion to load the upper bearings. I have nothing other than a notion at the moment but I think there might be something viable in the idea.

I never really thought the 'problem' with the axial load on offset skate bearings was more than notional. As I have said, I've seen so many videos of obviously working routers using that bearing design (including yours) that I couldn't believe that there was a particular problem with the design.

I guess I find the notion of avoiding the axial loading some how more aesthetically pleasing (if that makes sense to anyone, and I wouldn't be surprised if it doesn't )

I'll do do some more sketches over the next few days and see what people think.

[GPT]

GPT:
You clearly are a recidivist tinkerer
Z.

I had to look up the meaning of 'recidivist'

I have to say my wife would agree

[louieatienza]

I had to look up the meaning of 'recidivist'

I have to say my wife would agree

There is an evil connotation to the word that conjours thoughts of seedy individuals!

[louieatienza]

I'm hoping that cutting all the framework for the torsion box on a table saw without adjusting the fence will mean that the skins, and hence the rails, will be parallel. Having said that, I'll have to actually put it together to see how parallel I can make it

I agree that some kind of more robust preload adjustment mechanism is probably in order. You mentioned the springiness inherent in the offset skate bearing arrangement which has set me to thinking about using actual stiff springs in some fashion to load the upper bearings. I have nothing other than a notion at the moment but I think there might be something viable in the idea.

I never really thought the 'problem' with the axial load on offset skate bearings was more than notional. As I have said, I've seen so many videos of obviously working routers using that bearing design (including yours) that I couldn't believe that there was a particular problem with the design.

I guess I find the notion of avoiding the axial loading some how more aesthetically pleasing (if that makes sense to anyone, and I wouldn't be surprised if it doesn't )

I'll do do some more sketches over the next few days and see what people think.

I think however, that "pre-loading" the "offset" or 90-degree bearing configuration puts a radial load on the bearings, and the whole idea is that the bearings are pre-loaded to overcome any "axial" loads that occur.

As for your bearing design, the idea is similar to the Rollon linear rails I use on my current build. Even with a tablesaw, it's not easy to get a truly parallel, square edge! I wonder which way would be the best for the springs, the upper or lower bearings?

[GPT]

I think however, that "pre-loading" the "offset" or 90-degree bearing configuration puts a radial load on the bearings, and the whole idea is that the bearings are pre-loaded to overcome any "axial" loads that occur.

As for your bearing design, the idea is similar to the Rollon linear rails I use on my current build. Even with a tablesaw, it's not easy to get a truly parallel, square edge! I wonder which way would be the best for the springs, the upper or lower bearings?

I can't say I'm up with the dynamics (or is this a statics problem?) to calculate how the reactive force of the rails prevents a twisting moment on the bearings in the offset design.

As to cutting parallel pieces on a table saw, did I point out that my design would hopefully compensate for my sloppy workmanship I know I am not a brilliant craftsman!

I guess the upper bearings would be the best candidate for springs as gravity provides a natural spring for the lower ones.

I imagine the only lifting force comes from the router bit plunging into the work piece, at least for simple level cuts. I'm sure complicated 3D profiles are a different matter.

As and aside, very few things I try and build ever work first time. Still, I'm not easily discouraged

[GPT]

Got to do some work over the weekend. My opportunities are generally limited to when my 4 1/2 year old is asleep.

It seemed to go OK except when I lay the top skin on I measured a 0.2 mm difference in height from one end to the other. That was using a rusty old vernier caliper but I expect it's in the right ballpark.

So, while I will attempt to correct the difference I expect I will focus on developing some kind of spring loaded mechanism to load the upper bearings.

I'm currently imagining having the upper bearing axis pass through the short side of a square tube lever, while the long side of the lever is pulled down by a spring. The sprung lever will be effectively horizontal with the spring to compensate for the tenths of millimeter variation between the top and bottom rail.

As I pointed out in my first post, the design has to compensate for my shoddy workmanship

Cheers

Gerard

[CarveOne]

I find that my own shoddy workmanship is beneficial, in that it teaches me to figure out a better way to do it next time - after buying a new shop tool just to correct the situation. Yay(!) for new shop tools.

CarveOne

[GPT]

Yay(!) for new shop tools.
CarveOne

There is definitely something to be said for having the right tools.

Louieatienza commented earlier that:

"Even with a table saw, it's not easy to get a truly parallel, square edge!"

Now I know a bad craftsman blames his tools, but I expect the borrowed Triton workbench and GMC saw I used weren't totally top of the line

Having said that, this is still the most accurate work I've managed ever and milling 0.2 mm or so off the spoil board shouldn't be an issue.

Just hope the inaccuracies don't stack up to the point where the machine doesn't work

[CarveOne]

There is definitely something to be said for having the right tools.

Louieatienza commented earlier that:

"Even with a table saw, it's not easy to get a truly parallel, square edge!"

Now I know a bad craftsman blames his tools, but I expect the borrowed Triton workbench and GMC saw I used weren't totally top of the line

Having said that, this is still the most accurate work I've managed ever and milling 0.2 mm or so off the spoil board shouldn't be an issue.

Just hope the inaccuracies don't stack up to the point where the machine doesn't work

Louie is right, but the higher end table saws can be very precisely aligned with the newer specialty gadgets that can indicate when the blade is parallel to the guide slots in a cast steel table top. I too struggled with no table saw, or low grade table saws most of my life. To do precision work easily, you need precision equipment. But precision work can be done without the precision equipment with a lot more effort and self training.

Build it the best you can and with as much precision as you can. Provide alignment adjustments wherever you think it may be needed. Align the machine as close as you can get it and then surface plane the table top to make it flat and square with the gantry. That will minimize how much needs to be removed with surface planing. Anything that doesn't work can be corrected with an upgrade. (My usual plan of attack when things don't work.)

CarveOne