What's new
What's new

CNC plasma cutter - Make Vs. Buy

peckham4

Aluminum
Joined
Feb 17, 2005
Location
Castle Hayne, NC 28429
I have a customer who has promised me 3-4 jobs a week for a CNC plasma cutter. This will just pay for a plasma setup. Also I know if I had the setup I could sell other jobs for it, which at that time I will start making a profit. So now I must decide on how to go about getting a CNC plasma setup, and this is where I would like to hear your thoughts. I want something that will accept a full sheet, so 4 x 8 or 5 x 10 is the size I am looking at.

My first option would be to buy a complete setup. This would cost me more but would be less work for me and I could get up and running faster. So I guess here I am asking if anyone has any experience with off the shelf units.

My other option would be to build my own setup. Now keep in mind I build custom machines. I have already built X, Y gantries for customers before. I have also developed my own software for converting .dxf files to X, Y motion. So software is not a problem. However I have never built a plasma table before. While it seams pretty simple I always like to ask how other have done it. No reason to reinvent the wheel here.

And last how does plasma work with painted surfaces. I will be doing a lot of cutting on pre painted panels. While the cut edges will be covered I can’t burn the paint to far back from the cut or it will show. I have heard of cutting panels in water to help pull the heat out to keep from burning off to much paint. Is this standard practice? Do I need to design my table so I can fill it with water to just above the part?

Pete
 
I will not be able to help you much here but maybe just a little.
First off, this topic has been kicked around over on the Hobart Weld Talk forum.
Secondly, if you have experience building things similar to this, I believe you very well could build one. You might order a video of the plasmacam being used. I have not heard many good things about this product but it may point you in a better direction as far as building your own.
Lastly, I don't know your budget, but for full size sheets and painted surfaces, a waterjet (flojet) machine might be more to your needs. It will also give you the advantage of cutting many different materials other than metals. Personally, I would love to have a waterjet but I could never get it paid for.
Sorry if I was not able to help.--Grant
 
I had thought about water jet but I also just assumed it would be out of my price range. But to tell you the truth I have no idea what one cost. I do know laser is way high. I wonder if anyone knows about how much water jets go for?

As far as my budget goes its like this. I would like for setup to pay for itself in less then 6 months. I am planning on spending 3k on the plasma cutter and machine tip. And then a another 1k on each stepper and drive. And then about another 1-2K on table and linear motion. I have most of the material to build the table in house. So I will save some money there, and use up some of my stock material at the same time.
 
Waterjet machines, new, run from about $75k up to 2 or 3 hundred. My local waterjet shop runs a 5x10 machine from Chukar that I think they paid $150 grand for. The cheapest one I have ever seen is the Knuth, made in germany, for about 60 grand. Its funny, but waterjets are one of those things that no one will tell you the price of on their website- if you have to ask, you cant afford it, I guess.
And consumables are pretty expensive, too. I am not sure, but I dont think you can run the abrasive thru more than once.

New plasma machines that will take a 4x8 sheet cost about $25k to $50k.

So if you can really build one for 3 or 4 grand, less the power supply, you ought to go for it.
You will really want a third axis- torch up and down. You shoud put in a voltage sensing circuit to keep the torch automatically adjusted a constant distance from the material. My C&G machine, which I bought new, has 6" of vertical travel on the torch, and there are times, when, in cutting one sheet, it might move almost that far.
Steel will warp as you cut it, and move, and then when you come back around, its not in the same plane it was.

I have a water table below mine- the water is about an inch or so below the metal being cut- this catches a lot of sparks, swarf, and dirt. It cuts down on smoke, too, because the stuff goes out faster.

The really big professional machines sometimes do cut underwater- just by a 1/4" or so. I had some large aluminum parts cut- I didnt want to cut 40 sheets of 4x10 in my shop- aluminum puts out clouds of white smoke- and the shop I had do it used a 20' x 40' underwater setup- with about a 300amp power supply. Huge thing, the size of a reefer on its side. The torch looked like a regular machine torch, but it could have been special for underwater usage.

As far as paint- you will have to try it and see. The HAZ is pretty small, but paint will burn or discolor for 1/8" to 1/4" or so, depending on material, paint, and thickness of metal. The thicker metals take more heat, and a slower speed, so more paint burns. Thin stuff, like 16 ga, you can run flat out- I run it at 100ipm or so, but my machine is older, and my power supply isnt huge- with a bigger 200amp or so power supply, I have heard of twice that speed.
 
Ries,

Thanks for the great post. You answered a lot of questions for me. I was debating if I would need a 3rd axes, thinking I could always add it later. And I really had not figured out how to control it anyways. But you make some good points and using the arc voltage to drive the Z-axes will be easer than I had planed.

I am moving into a larger building in a few months. And I want to begin construction on this as soon as I get in there. So I have some time to work out my design. I will keep you posted on how it goes.

Pete
 
"New plasma machines that will take a 4x8 sheet cost about $25k to $50k."

They use to cost that, now they run around $10,000 or less for the "hobby" style machines. By hobby I dont mean they are inaccurate, I mean you cant run the forklift into the side of them and expect to get a very long life out of the machine. TorchMate.com is a pretty good machine and excellent software, practicalcnc.com has another good machine for a little less money. Plasmacam has a small 4x4 machine that has great software and decent machine, but I don't recomend them b/c customer service is **** to not existant, all the machine components are machine specific and you have to go to them for parts( the rest of the small cnc guys use off the shelf cnc parts that are easy to find). I would recommend going to cnczone.com and browse through the homemade cnc plasma table section. If I was going to let an idiot run a cnc plasma machine I would go with an used esab machine (about in the 30 thousand range used) they are built like tanks and can handle several sheets at once and has multible torch heads. If you was going to run it, I would go with one of the cheaper machines. The local school just bought a torchmate III with hypertherm 1250 and the total cost was about 15,000 with all bells and whistles, that includes ATC (automatic height control) On thick stuff the height control isn't too big of a deal, but anything 1/4 or less its almost a have to have option unless you want to babysit it. Oh yeagh I forgot about dynatorch.com, I have heard excellent things about them but havent run there machine yet. Don't let noone tell you cant build a good machine, the plasma cutting process is only accurate to at best .030 so the machine doesn't have to be rocket science. Look at the other small cnc guys and copy there designs. If you decide to buy a machine and things don't pan out you can sell the machines at about 60-75% of what you paid.
Chris
 
You can build a good machine today for a reasonable cost. Mach2/3 has a mode for plasma control, and there's a THC board available for less than $500 which integrates directly with Mach. Because its a non-contact process, relatively small motors, etc can be used, as compared to whats used on mills and other edge cutting machines.

IMO, when you look closely at what you get for the dollar, the hobby class machines like the shopbot, plasmacam, torchmate, etc are all grossly overpriced. A universal characteristic is no attention whatsoever to access for sheet loading. Half-assed design gives you a machine with a corral around the area you need access to. The plasmacam is a collection of brake formed sheetmetal guided by a few $2 ball bearings. Roughly $8000 without the PC, power supply, THC, or torch. $15,000 to Shopbot gets you a fully equipped pile of angle iron, more $2 bearings as guides, and proprietary operating code in the plasmacam tradition.

If you want something that doesn't look like it was engineered with a stick in the dirt, and built under a tree, either build it yourself or prepare to spend at least $30K for a decent machine.
 
something that doesn't look like it was engineered with a stick in the dirt
For your possible humor, many years ago I bought one of the first Shop Bots...like maybe the 4th one they ever made...for something like $1,200. This with the express idea of turning it into a CNC plasma machine and selling them as a product.

Never got around to assembling it, making a table, etc. and sold it to a friend of mine who put it together and used it as a CNC router for light machining of aluminum. Being one of the first machines there were numerous glitches with the software and such. He gave up on it a few years ago and now it's gathering dust. So, if anyone wants one of the first Shop Bots dirt cheap, contact me ;)
 
Ok it looks like I will be building my own setup. The professional machines at 20-30K are a little out of my budget. The lower priced machines appear to be over priced junk. For example one outfit wants 7K for 2 stepper motors, a couple racks and pinions and a couple pieces of extruded aluminum. Also they use cam followers on soft cold roll plate for linear motion. I think I can do better then that.

I also see this as being a great sales tool. I am in the custom machine building business and it would be nice to have a machine on my floor that customers could see. Right now I only have pictures of past machines that I have built. This simple machine would not only make money as a plasma cutter but would demonstrate our ability to build quality machines with motion and controls.

I do have a few more questions? It looks like most machines that I have looked at are using rack and pinion setups for the motion. Is this how everyone is doing it? Does anyone use acme screws or ball screws? If plasma is only good to .030 then rack and pinion should be fine. It will be faster and a lot cheaper to build too.

And what about motor size? The motors only have to act against the forces of inertia from moving the gantry. It looks like 3.25 ipm cutting and 5.00 ipm rapid traverse is what most machines are capable of. Does this sound right? I won’t know what the gantry weighs until I get further into the design. Can some one tell me the specs on their motors? This would give me a starting point anyways.

Thanks again
Pete
 
Pete, if you find you can build a professional grade CNC plasma with "open" software, and sell it for hobbist type prices at decent profit, you might consider getting into the CNC plasma machine business. I only mention that because if of interest, that might influence your design and parts selection on the first machine.
 
There are two levels of accuracy in cnc plasma- and they are determined by the torch/power supply, which leads this particular wagon.
With a normal, $3000 or so plasma power supply, and a $500 machine torch, you can only do so well- the torch cuts a cone shaped, rather than straight line, cut. This is accentuated when going around corners- there will always be a "good" side and a "bad" side. The thicker the material, the more noticeable this is, so plasma accuracy falls with material thickness. Thats why its a cheap solution for low tolerance parts, but picky stuff is either cut with a laser, or a new generation "smart" water jet machine, which actually swivels the head angle to put all the slope on the scrap side of the cut.
Next step up is a so called "high definition" power supply- suddenly, you are talking real money. Hypertherm doesnt even quote prices on websites or price lists- you gotta ask. But I have heard $20,000 to $30,000 and up, depending on amperage. Just for the power supply- no cnc or table here.
http://www.hypertherm.com/mech_systems/ht4400.htm

So if you arent spending the money on a hi-def power supply, rack and pinon is perfectly good- because the restriction in accuracy is elsewhere.

I have a commercial machine- a C&G I bought new in 92. It has rack and pinon, and pretty small motors, and it is as accurate as my plasma will allow me to be, plus its simple, easy to maintain, and pretty bulletproof.

As far as cutting speeds, I am assuming you meant hundred inches per minute- and again, you need the high priced power supply to approach those high speeds.
With a $2000 to $3000 off the shelf power supply, your cutting speeds are gonna be more like-
1/4" plate- 25 to 35 ipm
1/8"- 40 ipm to maybe 60ipm
gage material- 75ipm up to maybe 120 ipm or so.
rapids, as fast as you want, but they dont make much difference, on a manually loaded machine- you gotta manually pull skeletons and drops out, reload sheets- it just wont save you that much time if your rapid is 400ipm instead of 200ipm.

Speed is very dependant on how new your tip is- as they wear, speed will need to slow down. Speed is also affected by the complexity of pattern- it can go a lot faster on a straightaway than in complex curves- you go too slow around the curvies, and the metal gets hot enough to weld itself back together after cutting, too fast and you miss some corners, making it tough to break the parts out.

I have seen 1/4" plate curl up 6" into the air as its being cut- depends on profile- if you are cutting a 1/4" wide strip off the edge of big plate, expect it curl up like a lathe chip- this is why height control is essential.

If you can really put together a decent machine for 10 grand or so, the marketplace is gonna be at your feet- because the "hobby" machines are a joke. And real ones really do start at 30k. In industry, they still, stupidly, build them all to take Burny computer heads, which start at 10 grand just for the hardware (no software included).

I routinely cut 4x10's of 3/16"- not big, by fab shop standards. But a plasmacam would crumple like a pop can when you throw one of those on with the forklift. We had a job a couple of years ago where we had to weld two together, before cutting- 4' x 16'. then you need to slide it around a bit on the table, to get registration right- You need a sturdy machine, and those things aint it.

I recommend a relatively light, strong, square tube framework, with an open bay where the cutting takes place, and then a separate, 4' x 8' x about 6" deep water table. If the cutter is separate from the water table, your customers can build their own water table, whatever size they want- 4', 8' or custom.
Think about fume extraction- lots of smoke, and its yucky. The best machines have built in fume extraction, but that makes the base heavy, big, and expensive.
But it wouldnt hurt to add a lightweight 6" or 8" dust hose right to the torch carriage, and hook it up to a smoghog or a torit.
 
I modified an old Linde multi head burn table.
Added conversion software to convert from autocad to cnc. Added a portable plasma torch that was not being used, just to try it out. That was 3 years ago. But I do want to buy a proper plasma table from Controlled Automation. We have 2 of their machines (drill-line and angle-line)We have other machines in the shop and their machines are far superior. Personally, I would start small and cheap and slowly grow.
 








 
Back
Top