[hpladds]

The members of a Hacker/Maker Space, here in Atlanta, GA (Freeside Atlanta) have collaboratively designed a CNC Router. We post the design seeking input from this august crew.

This design extensively uses 2" x 2" (1/4" wall thickness) square tubing. Most uniquely, we mill the surfaces of the square tube and use it as Y axis guides/rails.

Design objectives:

1) Cutting surface of 4' x 4'

2) Sturdy -- able to withstand the unforeseeable abuse of a Maker/Hacker space environment: collisions with runaway autonomous pogo sticks and/or folk lifts, keg bowling mishaps, etc.

3) Simple and self-evident mechanical design for ease of maintenance and repair.

4) Low maintenance.

5) Keep the use of proprietary components to a minimum -- allow for several of the parts to be machined by the builder.

6) Achieve tolerances that allow (at minimum) for wood furniture part manufacture.

7) Keep the design plans Open Source.

8) Exchanging the router for a plasma torch is contemplated in the gearing/motor selection, but is not expressly included in the design.

Construction:

The table base, Y axis guide rails, and the broad expanses of the gantry will be constructed of 2" x 2" (1/4" wall thickness) steel square tube.

Because we have access to a mill, we will mill surfaces flat and use the square tube as Y axis guides.

Because we have access to a water jet, we will construct much of the gantry's frame fron 1/2" aluminum plate.

Drive systems:

Rack will be mounted to the underside of the Y axes' rails, and Ehren's (CNCRouterParts) "PRO Rack and Pinion Drive, Nema 34" will be fixed to the gantry.

The X and Y axis will use acme screw drive systems.

[lazorus]

Good job on the BOM but i dont see a power supply. Instead of using a large PS for my last build, I used two smaller ones (48v). I hooked one of them to the Z and A axis, the other to the Y and X axis. They work great. The stepper drives I used (four of them) were the DQ860MA. i used to use a Gecko G540 but both it and the steppers got SCREAMING hot. I would have to go "offline" with Mach3 to cool them down every so often.

I also used the CNC routerparts gear. Good gear they sell.

I only have about 30 hours on my setup, YMMV.

[Trotline]

Does the mill you have access to have a table travel of over 4'? That's a pretty big mill. If not, then the plan to mill the tubing flat is in trouble.

Luke

[hpladds]

Does the mill you have access to have a table travel of over 4'? That's a pretty big mill. If not, then the plan to mill the tubing flat is in trouble.

Luke

Well you've hit upon one of my greatest concerns. Some feel that the tubing could be milled in ~18" sections. I have my doubts -- they'll be ridges every 18". Is that your concern as well?

I did get a quote from a local shop that promised to do the pieces in two sections 62.75" / 2 = 32.375. So a small hiccup (ridge) will be left at @ 32.375". The machinist believes that the ridge can be "filed" so as to be pragmatically unnoticeable. It adds about $150.00 to the build.

My other concern: welding the table and warping the tubing. We built in "adjustment plates" in order to shim where the rails meet the legs. That might "take some doing" as well.

[hpladds]

Good job on the BOM but i dont see a power supply.
Instead of using a large PS for my last build, I used two smaller ones (48v). I hooked one of them to the Z and A axis, the other to the Y and X axis. They work great. The stepper drives I used (four of them) were the DQ860MA. i used to use a Gecko G540 but both it and the steppers got SCREAMING hot. I would have to go "offline" with Mach3 to cool them down every so often.


I also used the CNC routerparts gear. Good gear they sell.

I only have about 30 hours on my setup, YMMV.

Yea, we're still mulling the electronics.

Hmmm... any theory as to why they got so hot? Are we talking two separate builds here: a prior one with Geckos, and another machine built with the DQ860MAs? Or did switching out the drives of the same machine cause the heat issue to go away?

Thanks for the reply!

[Trotline]

Ridges can, as you were told, be simply filed or sanded down. They're not really the problem. The problem is that none of the sections is very likely to end up co-planar. Oh, they'll look good by eye, and they'll be close, but close is rarely good enough. The welding / warping concern is valid. A good, skilled welder can minimize it, but it's something that you'll end up dealing with.

There are a lot of pitfalls in designing these machines. They're deceptively simple. Given that there are a number of proven designs out there, I'm not sure that you'll be doing yourselves a favor by not building one of those.

Luke

[hpladds]

Video of the assembly here: Assembly - YouTube

[lazorus]

AFAIK, the G540's just run hot. And yes, two seperate builds. The prior machine had the G540 on it. My current machine has DQ860MA's and NEMA 34's. I did a run yesterday, 5 hours non stop. The DQ860MA's were warm. Probably 15 Deg C. above ambient. The steppers were about the same.

[hpladds]

Ridges can, as you were told, be simply filed or sanded down. They're not really the problem. The problem is that none of the sections is very likely to end up co-planar. Oh, they'll look good by eye, and they'll be close, but close is rarely good enough. The welding / warping concern is valid. A good, skilled welder can minimize it, but it's something that you'll end up dealing with.

Thanks for the reply Luke. We do have a couple of experienced welders on the team, so we're pretty confident that we can "minimize" the warping issue. We designed into the plans "adjustment plates" which will allow us to shim the rails. So "fingers-crossed" that issue is addressed.

However, this is the first we've learned of the "co-planar" issue. This issue seems truly viable to me, if we mill the entire surface in one "pass," would that remove the co-planar issue? BTW, we plan to mill three side of the rail. Does this raise any other issues?

There are a lot of pitfalls in designing these machines. They're deceptively simple. Given that there are a number of proven designs out there, I'm not sure that you'll be doing yourselves a favor by not building one of those.

Luke

We reviewed a lot of plans, and found them to be slightly off for our purposes. In particular, we didn't want to use 80-20 -- preferring the durability, cost and weight (as a base for the machine) of steel. But your point is well-taken and has vexed us. In particular, the MechMate nearly won us over. Buying a RedSail machine seemed to be a very simple solution as well.

But hey, we're a Makers space. We want to make original stuff and share it with folks; however, we have no desire to make stuff that doesn't work very well.

I attached an animation of the assembly process in my initial post, but it doesn't seem to have posted. I'll post a Youtube link to it. If you have the time, perhaps you might notice some other red flags. As discouraging as it can be to receive bad news via solid criticism, the laws of physics are much tougher critic.

Thanks Again

[Trotline]

If you can mill the entire length of the tube in one pass, then it'll work fine. As I mentioned before, though, that'll take a very large machine to do. The issue with doing it in multiple passes is that removing the tube from the table and re-setting it is very likely to lose the original line. You'd end up with a dogleg, either upward or downward, and likely tilted to the side as well at the line of demarcation. It'd be slight, if the machinist takes care with the setup, but you want it to be nonexistent.

It would be helpful to see more details of the design. The render in your first post is pretty small, and I don't want to comment on things that I really can't see.

Luke

[Trotline]

The 2x2 tubing is the only element of the design that really causes me any concern. It's small enough that you're likely to have some flex there, both in the lower rails, and the gantry. I'd seriously consider stepping up to 2x4, at the least, and maybe some truss structure to connect the two gantry beams. Steel is much more flexible than people generally realize, and a CNC router encounters considerable force acting on the cutting tool, in use.

Aside from that, and the milling issues we discussed earlier, it seems like a workable design.

Luke

[hpladds]

The 2x2 tubing is the only element of the design that really causes me any concern. It's small enough that you're likely to have some flex there, both in the lower rails, and the gantry.

I'd seriously consider stepping up to 2x4, at the least, and maybe some truss structure to connect the two gantry beams. Steel is much more flexible than people generally realize, and a CNC router encounters considerable force acting on the cutting tool, in use.

Aside from that, and the milling issues we discussed earlier, it seems like a workable design.

Luke

Thanks again Luke -- The flex also was a big concern for us. We're not certain how to reinforce the table base (just yet), but we expect to add bracing after it is built when we'll have a better "hands-on" understanding of where and how to do the reinforcing.

We also considered 2" x 4" tubing for the table; probably will include it now -- doesn't seem to add too much $$. Thanks Again.

Cheers

[nukwaste]

I agree with the assessment that the design needs strengthening up across the long axis? 2 by 4 tubing or an thin I beam section might be a better choice.

I noted that you are not using Thomson ball-screws and ways on the Y axis. From the drawing, I don't see a method to handle flex in the X direction across the Y axis. It would seem that the as-designed Y axis frame members would flex up and down when the bridge traversed back to front. Having a bar stock way with a bearing riding on it would give a shimmable set of screws to adjust any slop out of the system.

By the way, shimming is your friend here. The rails WILL need shiming no matter what you do. What I have done in the past is used a piece of steel wire strung from each end of of the frame set exactly 2 inches above the way. We would then use a 2" gauge block to illustrate how far to shim across the X / Y / Z travel extents. Its old fashioned but it works.

I have always used these. Round Rail Guides - Full Support - Std Block

I would also recommend large gussets on the frame members and cross pieces on the frame a few inches above the floor.