[brianMWParticipant]

Hi eveyone, I had a question about increasing the bed size. On page 14 of the instructions it says that useable area could be easily increased by extending the rails and bed. I was wondering if the same would be true for the x-axis? And, what do you think would be a safe size before it became unstable/inaccurate? Would 2′x3′ be pushing it?

My intentions for this machine are creating ADA-compatible signs with Romark, an acrylic product. The more bed area the better, although the z-axis only needs to travel at most a half inch if that solves any problems. Thanks for all the help in advance! This project is so awesome.

[ChrisKeymaster]

@brianMW, the Y-direction is the easiest extension; my guess is you’ll run into problems with stability if the X-beam is lengthened by more than 6-8 extra inches (and that’s a guess).

You could add some extra reinforcements to the X-beam assembly to prevent racking.

One of the issues to keep in mind though is how to fabricate a very long X-beam part. Our laser cutters max out at 32″, so that limits the size of parts.

Like many of the desired mods, it’s possible but will add complexity & new design issues to your build.

Please let us know if you do end up building a bigger gantry!

[brianMWParticipant]

Thanks for the quick reply. The laser cutter has a max area of 28″ by 49″, so I think I would be good there.

When you say racking, do you mean the gantry binding up on the y-axis rails? Sorry if the terminology is a little off–this is all a bit new to me.

[ChrisKeymaster]

That binding effect may happen — I think the most likely thing that would happen is that, as the two vertical braces move farther apart, there will be more leverage on one side or another depending on which end of the X-beam the machine happens to be closest to during milling.

Since there’s no mechanical connection between the Y-rails and the vertical braces (just gravity) there’s a chance that the bearings on the far end will slip while the bearings on the near end stay where they should.

There will also be more of a strain on the connection points where the vertical braces meet the X-beam, so that would probably need to be strengthened.

[DIYengineerParticipant]

I believe with a little thought and 3d modeling it would be quite easy to do. Many other CNC’s use rail like tubes that the entire thing is locked into place and slides on, and uses a gear and rack drive system or chain drive. If you get really fancy you can kill 2 birds with one stone and use a ball screw setup :) I guess the real limitations are your laser cutter dimensions as Chris said. Although a way around that could be making a longer piece out of shorter pieces, but more engineering would have to go into it to make it strong and remain straight while cutting.

[TaylorKeymaster]

All,

This project can absolutely be scaled up. I think that Chris is pointing out that it can’t be scaled simply (as in lengthening one or two pieces in Illustrator). That fact aside, @DIYe. is right – a little collaborative engineering is all that stands between us and a dope upgrade.

Let’s leverage the community! If someone wants to throw down a challenge, knock out some proposal images. Chris and I will be happy to comment on feasibility, and we can also advise on structural decisions once the proposal enters the 3D model stage.

[PhillipParticipant]

It looks like it’s been about four months since the last post, but the direction this thread was going would really be of use to me about now…

I’m thinking about building my own modified Lil’ CNC:

–the machine would be used in making components for architectural models and furniture prototypes

–it would likely be made of plywood (Aircraft plywood? Baltic birch?)

–the spindle would be some sort of heavier router with a 1/4″ or 1/2″ collet (Dewalt D26670?)

–there would need be a 12″ x 24″ x 4″ cutting envelope. I would LIKE to extend the z-axis to 6″, though.

–to connect to my laptop, an Arduino or some sort of USB-compatible controller would be used to drive the steppers

The laser cutter I have access to right now has a max. cutting area of 12″ x 24″, and I would rather not have to use any sort of laser cutting service. What I’m thinking is that the gantry pieces could be built up from two or more thinner plywood layers with interlocking pieces, all glued together. Here’s some diagrams (stretched proportions and not showing the t-bolt joints):

http://i48.tinypic.com/2m6jv2u.jpg

It would only be a solution for wood, since it would be difficult to get clear acrylic to look attractive when the glue marks are visible. Still, you can see that the joint seams would always be offset from each other to minimize the weak spots. Does anyone see any obvious problems with this technique, or anything else I mentioned?

[TaylorKeymaster]

@Phillip,

I like your laminating idea for strengthening elongated/subdivided panel pieces. You may also find that a longer gantry along the X axis will lead to racking problems under the current design. It seems that driving a rod from one end can potentially cause the end closest to the motor to start turning before the far end:

This is still a hunch (aka a projected problem), but we may need more significant design improvements to extend X. One idea could be to drive the X axis with an ACME screw placed under the table. This design drives from the center, which may eliminate the racking problem. This design can drive the tool or the material (as in this example):

We haven’t done much research on laptop control yet, but there do seem to be some solutions out in the wild. Bill Hastings made an initial post about the Aruduino grbl, maybe you could get in touch with him for further development.

[rmadamsParticipant]

We have been working on something similar- we wanted to be able to cut the machine on a 12×24″ lasercutter, so we have rearranged the parts to fit that size bed, and modified a couple of the longer parts to fit together in segments. Such segmentation is pretty much experimental right now, since we will not know how well it works until we have the whole machine together and functioning, but when the long parts are assembled, the offset and bolted joints appear to be as strong in a unit as the unsegmented ones.

Although the entire set of parts has been rearranged, the resulting sheets have not been optimized yet, to optimally pack and to remove unnecessary cuts. This was okay for us, since we have a lasercutter, and the cost of extra sheets of Masonite is pretty nominal. As soon as we verify that our approach works, we will go through the optimization process.

If others are interested, we can easily share the current draft 12×24 sheets, but they are likely to change as we test, verify and optimize them.

For more pictures of what they look like, you can check out this flickr page: http://www.flickr.com/photos/42433826@N00/4760445431/in/set-72157624295079547/

[ChrisKeymaster]

@rmadams, thanks for posting those pics!

We tried a similar split-construction at one point, and it did not seem strong enough. We abandoned it pretty quickly, mostly since we have ready access to a larger laser cutter. Please keep us posted as you make progress.

As for sheet optimization, I’d recommend checking out RhinoNest. It’s a great plugin for arranging laser-cut parts. You can download a demo, along with a demo version of Rhino (if you don’t have Rhino).

[munchymonsterParticipant]

Has anyone tried using thicker MDF and scaling the pieces up? I was looking into converting the files for cutting with a CNC Router. If I scaled the whole thing up I bet I could find someone with a 2′x4′ cutting area or even 4′x8′ cutting area pretty easily in the DIY CNC Router based machines to help cut it out. If I did that, the “stretched” pieces would still be cut from a single sheet with no seams to weaken the integrity of the beams. Do you think it would work? I’m thinking you could double the size of the drawing which should lead to about twice the useable cutting area. Or am I ignoring something obvious? (Quite possible I’m feeling a bit fried right now)

-MunchyMonster

[TaylorKeymaster]

@MM,

Your idea would work for basic structural connections, but would not work for hardware/parts of fixed size (like the bearings, for example). Thus, scaling would require that you maintain the original size of certain mounting points in the design. You could scale the hardware along with the panels, but this would be overkill in some situations (moving up to 1/4-20 bolts for the tab connections makes sense, but moving to 1″ rod on the slides would be unnecessary).

[Author]

Posts