Shopbot and AutoCAD

[curtiss]

As I am a bit new to all this, Will a shopbot follow an AutoCAD file such as a golf course map?

Will it follow a hatch pattern in the fairways?

What material might be good for a golf course map/ carving?
Anyone near Springfield Mo with a machine ??

Thanks

Curtis Spfd mo

[daniel]

Autocad is the first step Curtis. You use a Cad program to draw the files, in your case the golf course map. You then need a CAM program to generate the toolpath for the machine to cut. The tool path takes into account bit diameter and other things. For 2D no problem, you can do this easy. If your talking about a 3D topagraphical map(did I spell that right?) that can be more challengeing. Do you have the golf course maped out in 3D? That would be the first step towards a 3D carving.

[gerald_d]

Curtis, being proficient in AutoCad is a great start to being proficient with CNC. But it is not as simple as plugging your AutoCad file straight into the SB.

While inside AutoCad, you need to "Explode" polylines and hatches down to individual line elements. Then you need some converter software (some basic stuff is supplied with the SB) to change your AutoCad data into a file that the SB will read (.sbp file extension).

Looking at it very simplistically, if you started with a simple 2D AutoCad file, and you plan to engrave your lines onto plywood with a thin tip cutter, you could start cutting within 2 minutes of finishing the AutoCad file, using the "freeware" that comes with the machine.

But you will also find that the sequence of cuts will be VERY inefficient. If you drew the fairways in a random sequence, the SB will cut in that same "random" sequence. That is okay for a beginner or hobbyist, but you can move up to better CAD programs that improves this situation. Plus the CAD software will take care of things that you possibly havn't even started to think of yet. (Things like: what happens if 2 of your AutoCad lines don't join exactly? Do you want the cutter to stay down and force a "join", or do you want the cutter to pull out out of the job for a new plunge to re-start the new line?)

If you are imagining something in 3D, Then there is a LONG learning curve with EXPENSIVE software. The least of your problems will be the choice of material.

[curtiss]

thanks for the information, probably not into 3d as yet, as this would take quite a while

Do you change bits in relation to certain colors on the AutoCAD drawing, such as a thin bit for lettering? How do you keep the different lineweights/ bitwidths separate when cutting?? THanks again,
Curtis

[robtown]

Gerald,
can you expound on this a bit:

"While inside AutoCad, you need to "Explode" polylines and hatches down to individual line elements."

I've had exactly the opposite experience. I make everything as a polyline, I've had problems with cutting order and a few other issues using items that have been exploded.

Recently I proggramed dovetails for a project, I had the sockets worked out as a series of lines that move Y to a certain point, move X into the wood, then move BACK the exact way it came to it's starting point, then another Y move to the next socket, X move, X move back, Y move, etc... (follow me so far?)

I found that using an exploded ployline that the SB would try to lift out of the socket sometimes, rather than double back on the X axis to it's point of origin, befor the next Y move.

[gerald_d]

Hi Rob,

The conventional wisdom on this Forum is that Polylines need to be exploded. We actually never do that because we never draw in polylines (except for making boundaries, but that is for closed shapes only).

Basically, the SB's dxf converter will deconstruct polylines back to basic lines where it can. Sometimes it will not recognise some types of polylines. Hence the popular advice of exploding first and not sending polylines to the converter.

But, you do have a interesting point, of preventing a "z-up" at the end of a line segment, that polylines seem to be solving for you. This issue would apply to all "undercutters" like dovetails. Maybe the other guys who are more familiar with the SB's dxf converter can chip in here? (We don't have recent experience on the SB file converter)

[robtown]

As a matter of drawing discipline I always draw polylines or make shapes into polylines (to make editing and revisions easier).

Thru my own trial and error, I've found that things work better going into parts wizard , for my work, when I make all shapes into polys (generally speaking).

[robtown]

Blind dovetails on a 10 degree angle...
4063

[robtown]

Longer shot...
4064

[cnc_works]

Fascinating, Rob. I would be greatly interested in set-up, jigs, method of work, on this one! I definitely want to explore furniture joinery, i.e. box joints, dovetails, mortise & tenon, etc., on the ShopBot.

[jay]

Rob, I am interested in how you are doing the dovetails. Are you doing them on the end of the x axis with the wood projecting down to the floor so that the cuts are actually being done in the x direction, then moving in y to make the next dovetail?

[robtown]

Donn, for these I did the sockets in the side pcs laying flat on the table in a jig. The fingers were done by hand with a jig I cut out of 1/8" alum (actually 2 jigs, one for tops and one for bottoms).

Jay, I could've easily done what you suggested for the fingers, cut a slot in the table top and created a jig to clamp the pcs vertically (at a 10 degree slant) and used the same cutting file I used for the sockets.

I'd be happy to put together a little synopsis of some sort, and/or share the files with anybody interested...

[curtiss]

as I am close to Arkansas, I am still a bit confused. So to stop, change bits and router something else out at a special depth or width you do what... ?

How long does it typically take to run a fairly simple cutting path ??

any good web sites for laminate materials ??

Thanks again...

CDJ

[jay]

I am interested.

[richards]

Curtis,

Sometimes we let the idea of automation get in the way of solving a problem.

My primary business is computer programming. Whenever a new customer askes me whether I can write a program to automate his accounting system, production line, etc., I always ask him how he performs the job without a computer. In most cases, the computer just makes things faster and more efficient. If the job can't be done manually, adding a computer probably won't help.

Using that line of thought, the Shopbot, in a sense, just makes manual routing jobs more efficient. Every cut that the Shopbot makes can be made manually. Granted, you might need a ton of jigs and the patience of Job to do the job manually, but it can be done. How would you do it manually? What kinds of bits/cutters would you use? How many different depths do you need to cut?

When you know how you would route the piece without a CNC machine, you'll be well on your way to knowing how you could use the CNC machine to speed things up.

Mike

[don_ask]

Mike, you may have been responding to Curtis. But, you just changed my thinking process also. I have been struggling with a design, looking for a software solution. The solution is how I use the router by hand and then apply that process with the Shopbot. I already have all the software I need. I just needed another way to look at the problem.

Thanks

Don

"Even if you are on the right track, you'll get run over if you just sit there"
- Will Rogers

[gerald_d]

Mike, that was a great post!

[curtiss]

One also must remember,

"If you don't make dust, you eat dust...." but I think it applies to horses in the old west...

Curtis

[richards]

Curtis,

Let's revisit the golf course problem.

If I were to create an AutoCad drawing of a gold course, I would probably visualize the course in much the same way that someone making a topographical map might look at the problem. That is, I would 'map' the course as a series of layers, each representing a different elevation. The number of layers would depend on the ratio of the course's topographical height to the model that I wanted to build. For instance, if the height of the model that you are creating is three inches and the height of the golf course, from the bottom of the deepest water hazard to the top of the highest mound, is sixty feet, then the ratio would be 1:240 ((60 feet * 12 inches) / 3 inches). If you decided that you would make 240 layers, then each layer would represent three inches of actual elevation on the course; while with 120 layers, each layer would represent six inches of actual elevation, etc. (assuming that my math is correct).

Now returning to your requirement to use AutoCad, it would seem that drawing 100+ layers, with each layer containing one or polygons and with each polygon constructed of hundreds or even thousands of line segments, it would seem that the project would be tedious. Even after all of the layers were drawn, you would still have to render the drawing into 3-D, unless the cutter/bit/tool that you used gave you sufficient 'smoothness'.

With all of this in mind, and if I were trying to use a pre-existing AutoCad drawing to create a Model that could be cut on a Shopbot, I would probably start by looking for software that could render an AutoCad drawing containing hatch patterns, etc., into a 3-D file that could be converted to shopbot code.

Mike