Results 1 to 8 of 8
06-05-2004, 12:18 AM #1
I apologize if this is a stupid question or one that has been posed many times.
I am working with a plasma cutting table currently controlled by a photosensor (follows line drawings). For the sake of precision, consistency, etc. I would like to convert it to CNC. I have a PC and have access to much of the software I would need, but I'm curious about the process of conversion.
The current photosensor and controller have no make/mfg markings and so I can't find whitepapers about them, and I have no real way of knowing what kind of data transmission occurs and how to reproduce that with a computer. If it's just G-Code and I can hook it up via serial, I don't think it's really a big problem but I'm really not sure what I *should* do to make this a decent, relatively inexpensive conversion. I am just working with X/Y coordinates right now and extremely high tolerances are not my concern (considering the type of cut with the plasma tips used, precision is relatively moot anyway).
Any help, tips, etc would be appreciated, or just point to a FAQ/tutorial if the topic has been revisited a lot.
06-05-2004, 04:45 PM #2
There is a Yahoo group that deals with home made CNCs.
A company that you might look at is Ajax CNC.
06-08-2004, 11:12 AM #3
What information can you give me about the tracer system you have? I used to be in the business of making these machines, both the photo tracer and the CNC types. My company was C & G Syatems, now part of Thermadyne.If you can give me some info-, size, manufacturer, motor type, etc etc I may be of help.
06-08-2004, 02:26 PM #4
Thats what I love about this forum- the brain trust is amazing.
I actually have a C&G systems optical trace cutting machine. At the time I bought it, in the early 90's, I was told the only option for making it CNC was to buy a Burny control unit. In fact, the size of my control box is exactly the right size for a Burny to sit right on top.
But for what I want, Burny's are not very well designed, and unbelievably expensive. The basic unit that does squares and bolt patterns is something like 10grand, last I checked.
So if there is a sensible way to hook up an off the shelf PC to my C&G, and run some cad/cam software that doesnt cost as much per seat as a new Hyundai, I would like to hear about it.
06-08-2004, 02:29 PM #5
Thanks for the replies so far.
I am not in the area at the time so it may take a day or so before I can get the quick-n-dirty specs available (maybe even pics to help elaborate since I couldn't find a MFG/model in the tracer) for you. Thanks a bunch!
06-09-2004, 11:04 AM #6
While I have not been active in the cutting machine world for several years, the Burny 2.5 was a good choice as a retrofit for the C & G machines. The 2.5 had capabilities VERY similar to their Burny 3 and was somewhat less expensive. It could operate off the serial port for program input. It used encoders for feedback from the two axes and had cutter comp also, as I remember. If one could pick up a used unit (there should be a number on the market now), it would be a good choice as it was made for the specific cutting machine job. You might look on EBay and see if there are machines available.
The simplest PC based solution would use steppers instead of the dc servos standard on the C & G machine. I prefer the servos but steppers would be simpler to implement.
Because of the nasty hi-frequency noise from plasmas, some people choose a battery operated laptop for the controller on these machines.
This subject is still of considerable interest to me. If anyone would like to pursue it further you can reach me directly at [email protected] dot net. (Correct the obvious)
06-22-2004, 06:12 AM #7ashuford Guest
I haven't been on the forum for some time and do not want to interfer but you may want to know Mr. Leon Drake.
I bought a Torchmate CNC Plasma Cutter some two years ago. I purchased the optional torch height controller with this unit. We had some slight problems with the controller due to my Hypertherm 900 having a slightly different pierce voltage timing problem. Since Leon builds the height controller for Torchmate, he was the man I contacted. Leon drove from Gurley Alabama to Pleasant Hill, Louisiana and made the required modifications on my unit. It required a .5 second delay in a sensing command.
Leon has not gone to bigger and better things. He can be contacted at [email protected] His new partnership is offering servo-driven kits.
They don't come any better than Leon Drake. He has a personal touch with his customers that is almost unheard of these days.
08-01-2004, 08:11 PM #8
I've got an Airco burning machine with a Westinghouse optical tracer and want to convert it to CNC too. Here's what I've found out so far.
The motors on these machines usually have tach feedback instead of the encoders needed to retrofit a diy-cnc control. Depending on the motor configuration, the tachs can be replaced with encoders, the motors can be replaced with servos and encoders, or the motors can be replaced with steppers. Some may have servos with encoders, and if so, thats the ideal starting point, but I think thats a rare thing. The motors on mine have the tachs sorta permanently attached to the back end of the motors, so I'm probably going to use some small steppers in their place.
You wouldn't be using the PC based control to feed signals to the existing "box" but rather the PC would replace the box, and, in conjunction with some inexpensive drives and a power supply, the PC now controls all the motion.
Here's most of what you need to do this, along with approximate prices.
PC running Windows 2000 or XP $500 new
Mach2 control software. $150, and a fully functional trial version is available for download free from http://www.artofcnc.ca This software converts the pc to an all in software cnc control using step and direction signals to control up to 4 simultaneous axes of motion. If using plasma, mach2 has the ability to operate with a torch height control, and has plasma specific algorithms to control the motion thru corners where the lag of the cut causes beveling under normal conditions. The limits on the free trial version are in max operating speed and limit to about 1000 lines of code. The speed limit would prevent fast rapids but wouldnt affect normal cutting speeds at all. With the $150 you get free lifetime support and upgrades, downloadable from the above website.
breakout board about $125 this device optically isolates the outputs from the computer to the drives to prevent damage from back currents, etc and also has on-board solid state relays for activation of auxiliary functions such as spindle, coolant, plasma, oxygen valves, or whatever needs to be turned off and on.
2 geckodrive solid state drives. about $115 each for stepper drives and about $145 each for servo drives.
Power supply for the drives/motors probably $100 max since these are small motors.
Adding for incidentals like wiring, terminal boards, enclosure, etc the total will run less than $1500 even if both drive motors are replaced with new ones. As Ries pointed out previously, this amount would barely buy the box a commercial burning machine control come in.
One last thing.....several guys across a wide area (california to the uk) have worked together to produce this stuff, with the idea of it all integrating together easily. One guy in Michigan (i think) Tom Caudle, is an electronic engineer who makes a living running a CNC plasma table he built. He designed an electronic torch height control that works on arc voltage for plasma cutters. Tested it with a bunch of people for several months and put it on the market at about $500. When it was abvious Tom's control was going to be a real thing and not just talk, Art Fenerty, the developer of Mach2, added the ability to Mach2 so it can interface with and operate with the THC. The guy in the UK, Les Newell, is developing a super easy and functional cam program for sheet goods cutting, regardless of whether for router, flame, plasma, or whatever. Its in beta right now and he's encouraging folks to download and use it and provide feedback on any problems encountered. Its at http://www.sheetcam.com If a person is familiar with ANY cam program, they'll be able to produce code with sheetcam in about 15 minutes after its installed. He's got provisions for automatic lead ins, kerf comp (ever notice the direction of travel on a tracer shifts the kerf offset from one side of the line to the other?), dwell times, and so on. Best of all, its free to everyone who gets in on the beta. If anyone downloads and uses it, you need to check back to the site once a week or so for updates, since he's continually tweaking the features and performance.
All of this stuff I've mentioned comes with one very important thing.......access to the very people who develop it. They're available on the Yahoo forums, as are several hundred users, to answer any question about anything that pops up. And, unlike a lot of the control and software types we're used to dealing with, if there's a problem these guys all want to know so they can fix it. None of the typical "no one else has had that problem" or "our equipment is perfect so it must be your fault" stuff that we've all heard before. They just ask what you were doing when it happened so they can try to replicate and fix the problem.
I know things like this aren't for everyone, but its to the point where you don't have to be any electronic whiz to put one of these systems together. I'm in the zero to minus area of knowledge when it comes to electronics, but I've done one of these conversions on a stepper Bridgeport. Used the original motors and power supply but new microstep drives. Spent around $1300 including a new PC. Probably had less than 40 hours actual working time in the whole thing. It will now do helical interpolation and continuous contouring where before it was exact stop at the end of each block and only 2 axis circular interp. I plan to add a VFD, which Mach2 will also control, and then it will do thread milling and rigid tapping. As of right now, I figure I've saved in the neighborhood of $6K versus a package retrofit, and it's been a long time since I made 6 thou in 40 hours any other way. It wasn't difficult to do at all.