Physics undergrad decides to have some fun.

[Donnelly]

I came across this website after bumping into DIYEngineer on YouTube, and I've got to say this is a great resource. I've already read nearly a hundred pages of build threads so far and some of the stuff here is truly amazing. I'm amped to get working on my own.

So, I'm going to start off this thread with a few posts I've written up. They're mostly detailed background information and rambling, but it was helpful for me to type them to organize my thoughts and I like to keep detailed records of things.

As I get further into my ramblings and arrive at the point where I have questions, I've put the questions in bold, so they stick out.

If you have the time and patience to actually read my ramblings and give feedback, I'd greatly appreciate it!

[Donnelly]

Ok, so I am an undergraduate student at U. of MD, shooting for a triple degree in physics, math, and computer science. In hindsight, I should have done mechanical, electrical, and computer engineering, but whatever - that's not important. What is important is that I've got a mind that likes working with numbers, and in turn, likes designing things and working with tools.

Coincidentally, I've also been working as a personal trainer for over two years. I've got this going as my own LLC, with my website here.

A few months ago, I got the chance to begin oxyacetylene welding. That's right up my alley, so I jumped on the chance. And what did I start building? Fitness equipment, of course!

So, now I'm getting pretty good at this and would like to take this work from "hobbyist" to "light industrial." I'll be buying a MIG welder in the next week or two (looking at the Millermatic 211) and I'd like to get a plasma table running in the next few months.

Within the next 10-15 years, I could see this expanding into a chain of 5-8 gyms and a full-blown equipment company, but I'm getting ahead of myself here. Back to the topic at hand.

[Donnelly]

I'd mostly be working with 1/4" mild steel plate, batch producing brackets and other odds and ends for equipment. There'd also be a good amount of 1/8" plate, occasionally 1/2", maybe 1" on the rarest occasions. Accuracy to within 1/32" would probably be sufficient, but the more accurate the better. More important would be reducing the amount of manual work involved in measuring, cutting, and grinding. The ability to cut smooth shapes and bolt holes would also be big.

I'm into strength and conditioning type training, so here's one example of that sort of equipment that could benefit from a CNC plasma table. It's a set of speed sleds I built for for the football and track coaches at a local high school. Basically, you load it with some weight, strap it to the athlete, then have at it. Simple, but effective.



Currently, I've got to measure and draw all the cut lines by hand, then set up guide-rails to let me get straight cuts with my oxyacetylene torch. Then I've got to grind the cuts smooth, and don't even get me started on that... With a sufficiently large plasma cutting table, I could cut ten of these sleds out of a 4x8' sheet in very short order, with very little grinding afterwards, and that's the brunt of the work.

In fact, because the sleds are relatively simple and large, they'd actually benefit less than many other parts. There's better applications with equipment strongman and powerlifting, but nothing that I have photos of offhand.

[Donnelly]

Right now, I'm leaning towards making it purely a plasma table for the sake of simplicity. It seems like adding secondary functions to the table (eg. routing) greatly increases the requirements for rigidity, drive power, etc. Keeping it simple with plasma is nice, because there are no cutting foces from bits or torque on the gantry from mounting heavy spindles.

Still, I will keep these additional functions in mind as this build progresses, and hopefully I'll find some way to incorporate them:

1. Boring - The ability to put small (1/8 to 1/2") holes in sheet steel and square tube with accurate relative spacing would be awesome. However, if I can't get this, I could always do pierces at the proper positions to act as pilots and then ream them later on with a drill press. It simply wouldn't be ideal.

2. Vinyl/Stencil Cutting - For cutting vinyl stickers and spraypaint stencils for marking the equipment I produce. This doesn't seem like it'd be too hard, I simply have no idea how this sort of cutter is supposed to work.

3. Wood Routing/Aluminum Milling - If I could get both plasma cutting and light-duty milling in one package, I'd be in heaven. Wood routing would be for producing furniture (for myself). Aluminum milling would let me take on small jobs from my physics professors. I know most of them pretty well and they often mention needing some odd part for their lab work (not necessarily with ultra-tight tolerances), which they can't get from the campus machine shop because the shop is always overbooked and overpriced. However, this would probably be the hardest feature to add, and it's getting furthest away from my original goal of producing fitness equipment, so I'll probably drop this.

[Donnelly]

Practical Concerns:

1. Location - My machine shop is currently the garage of my rental home in College Park MD. The landlord is perfectly fine with my welding and whatnot, as long as I don't burn the house down. However, this isn't my house, and I'll have to move out eventually. I'll be graduating in 2.5 years, and where I end up really depends on where I get hired. It could be for some company local in Washington DC, or it could be all the way out in Washington state.

If my table is either large, heavy, or welded in a single piece, I might have to sell it after graduating. Is that a problem? Actually, with the people I'm friends with, I'm sure I could find a buyer, so that's no problem.

It would however, mean that this table would effectively be a test unit. That's not a bad thing either though, because that means when I come up with all my "If I were to do over again..." ideas, I can actually incorporate them when I do it over again.

If I built a smaller, transportable table, then I could take it with me when I move, but there'd still be a lot of hassle in disassembling it and reassembling it again, and if I buy my own house, I'm going to want to put the biggest, baddest tools I can afford into the garage, so I wouldn't even want to have any table that's so small it could be transported anyway.


2. Transportation of Raw Materials - I drive a Dodge hatchback at the moment. I can fit sheet steel up to 2x8', but no 4x8's, so is there really a point to having a 4x8' capacity table? Well, yes, because I can always rent a Home Depot truck for an hour or two or get a friend to help when I need to stock up on sheet steel. Futhermore, I'd like to trade my current vehicle for a truck in the nearby future. Before I graduate? Not sure about that, but maybe.


3. Purchasing of Raw Materials - My supplier does sheet steel in 4x8' pieces, so I'd need, at most, a 5x10' table. There's no need for a 6x12' or anything like that, and since I don't have a crane, there's no need for an x-axis extension to keep the gantry out of harm's way when bringing sheet steel down. I'll probably end up installing some sort of half-ramp with a roller on one end to help me load full 1/4" sheets manually. Thank goodness I lift weights, right?



Table Size:
I could probably get by with a 2x4' capacity table at this stage in the game, but in consideration of the above, I think I'll splurge and shoot for a bigger one. Regardless of the table size, you've got to buy all the electronics and the torch, which are the biggest costs. What's a few more feet of linear rail when you just shelled out $2500 for a torch?


Construction:

Assembly Type - If I were going for something small that I could take with me when I moved, I'd go for a bolt-together thing made out of 80-20 aluminum framing. However, for a bigger table, I'd go with a welded frame, which seems to be the popular choice. While I'm concerned about the heat causing warpage, that seems to be correctable (from Jellingson's and Millman's build threads). If I get good with MIG too, I should be able to minimize the warpage... At least, it'd be better than if I tried doing it with my oxyacetylene torch.

Perhaps the legs will be bolted on in some fashion, so they could be removed later to load the whole table on a flatbed, if/when it gets sold.

Water Table - Definitely going to want this to be a water table for cleanliness. Something starting with about a 3" depth on one side, then sloping down to a 4" depth on the other so I can drain it for the occasional cleaning. How and why do some people move the water to a secondary storage when the table isn't in use?

Other - Might consider appending a 4x2.5' area for welding on one end and storage underneath the table. Definitely give it some sort of leveling feet, and possibly wheels like what AndiKid is working on.

[Donnelly]

As a college student, I'd like to keep things easy on the wallet, but I can anticipate and accommodate for $2500-4000 in the table and electronics and about $2500 for the torch.

Fortunately, January is a good time for personal trainers such as myself, so I really hope to push work hard for the next few weeks to save up some money. Then I'll buy the table materials, shortly followed by the electronics, followed by the torch itself, so I don't need to have all the money all at once, but I can continue working as each new part comes in after the last.

Certain parts of the table, such as wheels or an area for welding, can be postponed or nixed for financial reasons.

[Donnelly]

Hypertherm gets a lot of recommendations, so I called and spoke with Jim Colt last week about which torch to get. Seems like the Powermax 45 would suit my purposes. I'll get it with the machine torch first and buy the hand torch lead later on (unless I can afford them together).

While getting only a hand torch could save some money, I don't want to spend a lot of time fiddling with the thing in a special mount whenever I change consumables.

[Donnelly]

4 Axis Stepper System, with DTHC - On Jim's recommendation I looked into the systems sold by Tom Caudle (might not be spelling that right; he posts here too, but I can't remember the username offhand). The 4 axis drives with stepper motors look sweet and I like the idea of having digital torch height control. I'd either go with the 620 oz-in. motors or the 300, depending on how heavy the gantry is planned to be (100lb and 50lb maximums, respectively, I believe).

I'm not going to even think about building my own drive system with parts from multiple companies. I don't have that kind of electronics expertise (should have done EE and CE!). A kit's the way to go.

I've seen a few people describe stepper motors as some sort of horrible, cheapskate option, but having seen what Millerman's oxyacetylene table can do, I'd be perfectly happy with those kind of results. I figure if I can just get the right gear ratios


Belt Drive for X and Y - I was considering a rack and pinion design originally, but I really didn't like the idea of having to deal with preloading the rack to reduce backlash. Getting all the parts arranged for the spring to pull the pinion and rack into meshing properly seemed so contrived and full of hassle. Not to mention that you've got to drill and tap the rack itself into the rack every couple inches. Doesn't sound too fun.

I've got to say that Fiero Addiction's belt drive system looked sexy. This will be driving a much longer system, but I may go for 3/4 or 1" wide belt with the lightest possible gantry to minimize stretching. So, the X direction (long axis) would be driven on both sides, as usual. The Y direction would have a single drive along the gantry.

Gear Reductions - No idea here. Need to plan a gantry (to find a weight) and do some math to find out what motors I'd like to use and how fast I'd like my rapids before picking a reduction size. I already know to keep the stepping rate as low as possible (depending on the desired resolution) to maximize torque.

Linear Motion - Borrowing from several of the desings I've seen, the vee rails look like a good option. I looked at the precision ground linear rails from McMaster, and damn they were expensive! (Over 30 cents per millimeter.) Those things would also be overkill with a belt drive and steppers. The vee's looked much better for this sort of purpose.

Revisiting the Multifunction Table Options - Now, with the combination of vee rails, a belt drive, and stepper motors, I'm looking at a machine that'll probably be within my desired 1/32" accuracy for plasma cutting. If I have vee rails on the top (to guide the gantry and Z axis mount) and the bottom (to stop either from being pushed off the top rail), then I should even be able to do boring. However, this doesn't seem to be beefy enough to handle that wood routing and aluminum milling I mentioned earlier, but like I also said, that's okay. It's simply not ideal.

Z Axis - Seems to be normally acme screw driven, since racks or ballscrews let the axis drop and crash when the motor is disengaged.

Do people normally buy a machined Z axis and bolt it onto the gantry, or do they make their own from scratch?

[Donnelly]

Through the university computer labs and a machinist I know on campus, I've already got access to AutoCAD (which I've used a decent amount) and SolidWorks (haven't used), which seem to be the two best CAD programs out there. So, questions:

- How do I take what I've done with AutoCAD already and turn it into toolpaths?
- What is CAM Works and what does it have to do with SolidWorks? (Remember, I could learn to use Solidworks, but I haven't yet.)
- What exactly is Mach 3 and what does it do? (It seems to be a required part of the motor driver kits from CandCNC.) Does Mach 3 come included with the CandCNC packages?

[Donnelly]

Ok, that's all for now. Thanks for putting up with my rambling, and even greater thanks for any comments you can give. I'm looking forward to watching the project unfold over the next few months. But for now, I've spent enough time thinking about this, and I've got to get ready for work.

[leeleatherwood]

Donnelly, the best tip I can offer is to use Solidworks combined with this website:

Free CAD Models, Free 3D Models from 80/20 Inc.

Input the model number of the 8020 piece you are looking for, download the file and insert the piece into a Solidworks assembly. Learn to use the Mate function of Solidworks and you can design yourself an 8020 machine in a few hours.

That website also has other models which will be helpfull such as NEMA motors, bearings, etc.

EDIT: You can also use that website to create custom length extrusions for you instead of modifying the part in Solidworks. Just use the custom length function, saves TONS of time!

[ger21]

- How do I take what I've done with AutoCAD already and turn it into toolpaths?

Typically with a CAM program. Although I wrote an AutoCAD macro that utilizes polylines and circles as the toolpaths, and exports the g-code for them.
For plasma, a lot of people like SheetCAM.

- What is CAM Works and what does it have to do with SolidWorks? (Remember, I could learn to use Solidworks, but I haven't yet.)

A CAM program integrated into Solidworks.

- What exactly is Mach 3 and what does it do?

Mach3 is a machine control program. You load g-code into it, and it controls your machine. It's $175