[BoxGods]

I am in the design phaze of my CNC router and have gotten as far as I can without educationg mtself further on some of the more basic elements and need some help even knowing what questions to ask at this point.

Machine specifications--known--

• 24" travel on X and Y with 6" travel on Z
• All Aluminum/steele construction for the machine
• Wood construction for the base (not shown )
• Nema 23 steppers
• Porter Cable Laminate trimmer
• Dual Slide Rails on all three axis.

As I understand it, Servos are better/more accurate, or report theur actual possition better then steppers, but on a well built machine steppers will do the job fine, and are a lot less expensive.

I will include some images of the preliminary design after my questions. I have searched through this form trying to answer as many as I could before bugging you guys.

1. Lead Screws vrs. Ball Screws. As I understand it, the ball screws are more accurate/repeatable. I was thinking 1/2-10 lead screws or the equivelant ball screws, but I am not sure how to incorporate them into the design (there is a detail of the slide in the images below).
2. How are the ends of the threadded rod where it exits the vertical (horizontal on Z of course) supports actually suspended? Are the ends machined to .25" dia so they ride on bearings pressed into the blocks?
3. Limit/Home switches. It seems they are needed. Is any one type better? What is the most econimical choice? Who do you recomend?

That should get me further along, and I don't want this post to be TOO long lol. Please note in the images bellow that the triangular gussets are not shown, but I have made provisions for them. Also, some og the screw/bolt holes are not drawn in yet either as its a concept sketch.



Above is an overview image. The rectangular tube used for the Axis slides is 2" x 6" aluminum with .125" walls. I like this choice because it makes for a very stiff structural material, is light, and the enclosed design will help limit the rails exposure to debris. The flat stock is .5" aluminum.



More detailed view of the motor mounts for the Porter Cable Laminate trimmer. the backing plate for the Z head is shown as .25" but I will likely use .5" also, and include some triangular gusset braces.





The 2 images above show some of the details of the box beams. You can see how the rails will mount. Also, the end blocks that are bolted on are shown at .5" but I have already upgraded them to .75" to accomodate 5/16" X .75" screws to secure them in place on the box beams. I show the threaded rod being supported by a bearing, but as stated above, I am not sure this is how it is supposed to be done.





Above I have shown how the carriages (no idea what the correct term is) without the box beam so you can get an idea of how they are assembled. They are 6" wide to provide good stability and tesist rotational forces (rocking). The center bearing is 2" Dia Derlin type (ala Widget Master's excellent machine) though I am not sure if you can use them with ball screws, or how the ball screws are mounted/incerted at all for that matter.





The last 2 images give you an idea of how the cariages/slides work with the slots and mounting in the box beam. As a novice to CNC Routers, the design seems very capable structurally, (to me anyway). I will have slight recesses or pockets for the stand off's machined into both the carriage blocks, and the .5" up rights to ensure everything bolts together true.

I appreciate any help you guys can offer, and I WILL be building this machine. I have access to a large CNC router for cutting the flat parts, and the slots in the box beams--all I am waiting for is to lock the design down. I also intend to share a full step by step building log with excellent images and video to pass along whatever I learn.

Thanks for any help, and Happy Holidays to all of you =)

Geno

[Oldmanandhistoy]

Hi Geno,

I am very impressed by your whole post; you have made every thing very clear and added excellent drawings. While I could answer some of your question I will leave that to the more qualified members. I am sure this is going to be a thread that is worth watching and will be doing so very closely. I wish you the very best developing and building your machine.

Best wishes,

John

[BoxGods]

Hi Geno,

I am very impressed by your whole post; you have made every thing very clear and added excellent drawings. While I could answer some of your question I will leave that to the more qualified members. I am sure this is going to be a thread that is worth watching and will be doing so very closely. I wish you the very best developing and building your machine.

Best wishes,

John

Thanks John =)

I think once you get bitten by the DIY CNC bug, this forum can be the difference between biting off more than you can handle, and a very rewarding project. I think about "my machine" virtually every free moment it seems lol.

G

[Oldmanandhistoy]

Thanks John =)

I think once you get bitten by the DIY CNC bug, this forum can be the difference between biting off more than you can handle, and a very rewarding project. I think about "my machine" virtually every free moment it seems lol.

G

Tell me about it; I think CNC virtually all my waking hours.

Back to your machine

One comment if you can go ball screws don’t look back.

Regards,

John

[Mike F]

Geno,

Just a point on the screws. If you use ballscrews, you would normally buy a corresponding nut to go with it, either with or without pre-load. Pre-load is a method of reducing backlash to virtually zero. Ballscrews are manufactured to allow the use of ball bearings, in the nut, to follow the track and thus reduce operating friction to a minimum. Only if you use a leadscrew (acme, all thread or whatever) will you want or need the delrin nut.

CAD work certainly looks good and should produce an excellent machine. Are you sure just 24" x 24" is going to be big enough? I hope you don't find the size to be too limiting, once it is built. You will be surprised at the number of jobs you will want to use it for

Mike

[spoiledbrat]

I love your design. My machine can be found in the gallery, and is also made of .5" alum plate. I would like to say that 1/2" aluminum plate and socket-head cap screw construction is very impressive.

For your limit switches, the truth is that the $5-$10 limit switch will suit your purposes nicely. Using the verify command in Mach II, I usually get less that 0.0007" error. This repeatability is all you need for woodworking.
Check this link:

http://www.allelectronics.com/cgi-bi...ER_LEVER_.html

these are not the exact switches I used, but you get the idea. I think I paid about 3 dollars each for mine, and the repeatability is rediculous.

As far as the ballscrew, I think you should go with the economical 0.625" x 5 tpi nook ballscrew, along with two ballnuts each screw. An anti-backlash coupling can be made with two ballnuts, spring washers (or bellville washers), and some aluminum or bronze. See my picture for this, they have worked great for two years, and at least 700 hours. My ballscrews were turned down to about .375" for the bearing block. One of the two bearing blocks has ang-contact bearings to eliminate axial play, and the other is a simple radial bearing.

A motor-standoff will be required to hold your motor at the end of the ballscrew, and allow for motor couplings to be used. These are slightly flexible to allow for slight misalignment. Please see pics of my machine to demonstrate what I am speaking of.

my gallery pics:
http://www.cnczone.com/gallery/showg...00&ppuser=7181

The coupler shown has two components. One is the aluminum block, drilled and tapped for mounting to a carriage, and also tapped 15/16-16 for the ballnut to screw into. Two .25" steel pins are pressed into it. A bronze rod is then tapped 15/16-16 to accept the other ballnut. These are put together, with belleville (spring) washers, and adjusted to the desired preload. Mine is (hopefully) set to about 25 lbs. I believe this overcomes all cutting forces I encounter, eliminating backlash. I will post a more detailed explanation of this device, if you are interested.

Good luck and have fun with your machine. I know it will turn out great.

[paulC]

Great drawings.
My comments are not meant to offend, just prompt thought.
While your design will limit the dust going in it will also trap the dust that does.
If the beam is supposed to act as a torsion box, by cutting slots the whole lenght will greatly reduce its strength.

On your questions: Ball screws have advantages in that the energy transfer is far more efficient and backlash is easier to tune out.
The ends of threads are normally turned to fit bearings.

Limit switches are nice to have but on a stepper machine such as this you can leave them until later. If your machine decides to run amuck it will simply stop when it hits something and the steppers will loose steps. No big deal.
You can use anything for limit switches most people use micro switches. More critical are home switches as these are used to set the 0,0 position if you decide to install them that is. These need to be acurate and be designed in such a way that they will not be moved out of alignment if the machine runs to its stops. There are ways to do this with micro switches or optical switches.
Paul

[BoxGods]

WOW, thanks Guys =)

I guess part of why I am having trouble with the concept of a ball screw is really just that I have never seen one, and when you look online, most are just text links with dimensions/specs. If they ever show a picture of it, they dont show it actually installed.

So, is it correct to say that the hole drilled into the carriage/slide is just drilled larger than the threaded rod (no contact) and the nuts are screwed to each end of the carriage?

Do ther protrude beyond the carriage (thus reducing travel) or are they recessed into it?

And Pauy, no worries as I feel feedback now reduces problems later. I was also worried about about removing the crud that does manage to get into the slots. I might add a slitted gasket material that parts as the stand off's pass, or may just remove the area between the two slots. My thinking is to try the machine and see how must it collects, and how hard it is to blow it out with the compressor, then if it is a problem, just cut the center portion away.

As for torsion along the long axis of the boxes, I d not think it will be an issue as each end is held motionless, and a lot of the professional machines use a "U" cross section anyway. I think it helps in stiffness a lot from end to end however. It seems over built anyway (overkill is the best kill =) when I run the stresses in Solid Works, though I am just learning thae program.

On switches, I had no idea they were so inexpensive, so I will certainly add them. As for the Home switch...what exactly does that do? I get the part about returning to a "home" location, but I am not clear when or what triggers this?

Thanks again for the help =)

G

[aspenelm]

hello,nice design. I don't see how you have access to bolt down the slides inside the tubing. It would be okay if you can get the slides with tapped holes on the rails, but this is uncommon I think. Also I don't see how you can bolt the carriage assembly on top of the bearing blocks in the fifth picture.

[BoxGods]

The guid rails are simple, they have threaded holes and are secured from the outside via screws. You can get them tapped, or save a few bucks and tap them yourself. Either way a non issue.

The blocks are mounted the same way, with the threads being in the carriage.

G

[paulC]

On home switches.
They are set up in your software and involked by gcode.
G28 or G28.1
They tell your machine to advance in a predetermined direction at a predetrimined speed until the home switch changes state. The machine is zeroed to this point or with an offset. Makes powering up your machine so much easier.
Paul

[BoxGods]

On home switches.
They are set up in your software and involked by gcode.
G28 or G28.1
They tell your machine to advance in a predetermined direction at a predetrimined speed until the home switch changes state. The machine is zeroed to this point or with an offset. Makes powering up your machine so much easier.
Paul

AHHH...I had it in my head that a home switch worked like a limit switch in that under some circumstance out in the cutting envelope the machine hit the switch and this sent the machine home. I just coul not figgure out what would make you WANT to do that lol.

Makes a lot more sense now that I know the home switch stops the drive at home, and doesnt send it there.

Is the home switch just on one axis...or is it really 2 home switches?

G

[paulC]

AHHH...I had it in my head that a home switch worked like a limit switch in that under some circumstance out in the cutting envelope the machine hit the switch and this sent the machine home. I just coul not figgure out what would make you WANT to do that lol.

Makes a lot more sense now that I know the home switch stops the drive at home, and doesnt send it there.

Is the home switch just on one axis...or is it really 2 home switches?

G

I think you can use your limit switches as home switches in some software but it gets messy. I have home switches on my x and y. ( You only have one per axis)
Z I'm a little crude with, in that i connect a clip lead to the routerbit and another to a plate I sit on the surface. These are connected to one of the limit switch inputs and as I jog down it stops. I then zero at this point remove the plate, jog down the thickness of the plate and zero again. I'm sure this could be done better but it works for me at the moment.
Paul

[BoxGods]

That actually sounds brilliant =)

Simple and ellegant work for me.

Thanks

[BoxGods]

An update.

I have drawn up a second design using a dual bar solid mount rail system that I like a lot. Just waiting to hear back from the manufactuer on pricing (to make sure they are not exceptionally espensive) before I post the new concept drawings.

The waiting part of this hobby sucks a bit, but every hobby is that way I guess :tired:

G

[wcarrothers1]

I zero by site on Z mostly or use a piece of paper jog down till I can't move it then back up. Although I also have dial MPG's which make moving everything around a ton easier then keyboard jogging. My pendant has 3 mpg's so I can move all directions which out having to switch around either.

My 2 X & Y limits are in parallel normaly closed. machine first homes Y which opens the switch. Software moves back off till the switch closes again then does the same in X. Took me a while to understand wiring the limits in series and such...

I'm also a servo user. First machine was servo so the controler I have is servo based so I continue to do servos. Although the thought of trying to figure out cutting speed based on the possiability of making sure I go slow enough not to move steps seems like it would be time consuming and result in a lot of waisted pieces. Cutting slightly bowed (or not 100% flat) pieces of ply wood is pretty common on my machine over 4x8' so depth of cut and pressure on the router varries becides some times ya hit hard spots in plywood that ya can't really see...

b.

[BoxGods]

I will upgrade after I learn how to use the machine an can afford it lol.

[paulC]

Good point on the normally closed switches. Should have mentioned this earlier for BoxGods. If you try to set up your switches as normally closed you wont get caught out with broken wires etc. You will know you have a problem when it first occurs.
Paul

I zero by site on Z mostly or use a piece of paper jog down till I can't move it then back up. Although I also have dial MPG's which make moving everything around a ton easier then keyboard jogging. My pendant has 3 mpg's so I can move all directions which out having to switch around either.

My 2 X & Y limits are in parallel normaly closed. machine first homes Y which opens the switch. Software moves back off till the switch closes again then does the same in X. Took me a while to understand wiring the limits in series and such...

I'm also a servo user. First machine was servo so the controler I have is servo based so I continue to do servos. Although the thought of trying to figure out cutting speed based on the possiability of making sure I go slow enough not to move steps seems like it would be time consuming and result in a lot of waisted pieces. Cutting slightly bowed (or not 100% flat) pieces of ply wood is pretty common on my machine over 4x8' so depth of cut and pressure on the router varries becides some times ya hit hard spots in plywood that ya can't really see...

b.