NEW CNC BUILD! Brilliant Minds needed within!

[diyengineer]

This is my first cnc mill i will be creating. I basically need a good laundry list of parts and suggestions from everyone so i can model it in solid works 2009.

The overall intent of this machine is to mill aluminum/copper plate with a spindle, and have a repeatable accuracy of .005".

My budget is roughly 3,000-4000 dollars (I already have Solidworks, Mastercam, Mach3, and a computer)

I have a decent shop at my disposal, with a wide range of electric, pneumatic and hand tools. My father is a Aircraft mechanic my trade so we have sheet metal tools, and a variety of everything. (Belt sander,Tig welder, drill press, air compressor,die grinders, you name it, we currently do NOT have a lathe or mill).

I would like a 24"x24"x5" workable area (roughly).

I would like to use Large (3"x6" 15series) 80/20 for the base as well as the gantry, to make it as rigid as possible, and to bolt together easily.

I would like to use DC Servos in a closed loop configuration.
(For increased accuracy).

I would like to buy the majority of all the parts new through a reputable online store. (Not ebay, So others willing to build this machine can easily purchase parts from a parts list supplied from various vendors. More power to them if the part is available on e-bay and they can acquire it for cheaper.)

All my design work as well as plans will be 100% open source for the entire community, i may even go as far as creating a website hosting my own server to distribute plans and parts list.

As far as ball screws and Linear bearings go i would love suggestions on how many to mount on each axis as well as what brands/models and sizes will work for me, remember i plan on machining aluminum so overkill is always better =)

If you suggest a part please try and link it to a store online that sells it to make my life a bit easier! Many of the ball screw/linear rail companies offer 3D models of there product which make life really easy for me because i can directly import them into solid works and don't have to reverse engineer 2D drawings to 3D.

That is about all i can think of for now. I would love to get an entire parts list down, find online store for the parts, 3D models of the parts, and start modeling. Overall design once we get that far will be offered to the forum so i can have as many people critique it as possible. There are far more brilliant minds on here that my own. Hopefully with help from all of you in the end we will have; Overall price, specs, parts list, vendor list, I will supply HOW to VIDEOS, solidworks files, etc.

Thanks for your help, hopefully we can come up with a machine that is amazing!

[diyengineer]

Last night i was modeling a bit in Solidworks 09'.
Base Parts list:
80/20 15 series 3060 (2x24" pieces, 2x30" pieces)
80/20 15 series 4417 18 hole 90 degree brackets (4 pieces+hardware)
80/20 15 series 2531 45 degree 12" brackets (4 pieces, with hardware)
Nook Industries NH-EA series linear rails and blocks (2 rails, 4 blocks)
1/4" 6061-T6 Aluminum top with 1/4" tapped hole pattern (3"x3" i believe).

Attached Pictures:
Render-2 Just a view showing the aluminum top.
Render-3 Top turned to glass instead of aluminum to show inside of table.
Render-4 Top turned to glass instead of aluminum to show inside of bracket.

I just wanted to get a feel and a idea of how its going to come together, parts will change as well as the overall size of the machine. I just want to mock up a initial machine in solid works to get the feel of the overall general gantry design, before i get too specific.

I was going to mount the linear rails on the sides of the base, but by placing them on table top the load is distributed symmetrically on the rails and bearings, overall easier to align them parallel to one another, and the force is distributed straight down through the plate compared to being mounted to a flimsy machined down fin, between the T slots (via the side). Disadvantage to this design is having to make the table a bit larger, because by top mounting the rails i loose working space. I believe for overall accuracy top mounting the rails would be the best route to go, and i don't mind making the table a bit larger to accommodate them.

I would like to use a dual ball screw setup so i get absolutely no binding, and the force and motion will be distributed evenly to the moving gantry. I may make the table wider and mount a single ball screw on the outside of each linear table rail. Then i can mount the Servo/Stepper in the rear of the table bolting small gears to the servo/stepper and both ball screws. Then link the entire system together using a belt, chain, or some other form of mechanical linkage (whichever is most accurate).

Tonight i will try and accomplish placing ball screws on the table and possibly making a rear stepper/servo mounting perch.

I would love for any kind of positive input! This will be my first cnc mill and im trying to do it right! Thanks!

[Bartuss1]

This will be the sliding bridge, or sliding table?
in a small structure to better examine the movable table

[diyengineer]

The Table will be fixed, and the gantry will move. Unless i should do it the other way, what do you think?

[Bartuss1]

the movable table 3 ballscrews just enough
For example

[diyengineer]

WOW! That looks way more accurate as well as simple! I did not even think of doing it like that. You pretty much saved me from making a big mistake.

Will these linear rails work?
http://www.glacern.com/sbr

[Bartuss1]

Linear ball bearing block are less accurate, but cheaper

[diyengineer]

Linear ball bearing block are less accurate, but cheaper

If you were me what would you use on all 3 axis for ball screws and rails?

[Bartuss1]

worst rail is better than the best roller,
the bolts just

[packrat]

Hi, now this is just my take on doing something like this, YMMV. The first thing I would(am) do is figure out what you are going to use for a spindle.

Why? because that will tell you how much bigger your 'X' axis will need to be to get the amount of 'workable area' you want. Why? Because your linear rails and screws will need to be longer than your actual movement. And without your basic "z" axis size how will you know.

While a CAD program is great, the best thing you can get for something like this is some poster board. I know, blasphemy on a CNC-CAD-CAM site!! But after you get a starting design, if you draw to full size on poster board and cut it out, you will get a better feel and possibly even find out things that won't work. Old school for sure, but still works as good today as it did back then.

Sorry to post so much in your thread, cary

[Bartuss1]

Engineer usually has a spatial imagination and AutoCAD.
What would go back to prehistoric times and cut out templates?
old school for dinosaurs

[diyengineer]

Well overall update: I have come to the conclusion that i almost made a huge mistake that i'm sure lots of people overlook. I was going to make the entire gantry move but based on what I have seen and been told fixing the gantry and having the table move adds an enormous amount of rigidity, which hopefully will add to overall accuracy.

My new debate is using linear rod rails with ball bearing blocks, or Hiwin rails and blocks.

Since i need 2 rails for each axis (6 total) and 2 blocks per rail (12 blocks) the hiwins make a huge dent in the wallet where as the Glacern machine tools SBR series is much lighter on the wallet and appear to be fully supported, and extremely well manufactured.

http://www.automation4less.com/linear.htm <===Hiwin rails
http://www.glacern.com/sbr <===Glacern supported rails.

any information regarding accuracy between the two would be appreciated, i dont plan to load anything heavy onto the table, maybe a aluminum plate that weighs 50 pounds if that, and if so rarely. Accuracy is what im more concered about, and if i can get by with the less expensive rail then that would make me happy.