Plasma table project build (Perpetual)

admin

Was given some free buttress thread rod and nuts (stainless)Have lots of slo syn stepper motors.The arc-pak project is not finished, and the Miller wire feeder fix up not even started. Naturally it's time to start another project.A CNC xyz plasma table  Attached ImagesBest to use common sense before doing anything shown above.
Reply:There is certainly nothing wrong with starting another project, but you should find another application for those buttress threaded rod. Transferring power in both directions equally would require acme thread as an economy drive, or ball screw for more precision. Where did you get all of the steppers? Have you looked into controllers yet? Good luck!
Reply:Oops my bad it is acme. Thanks for the correction.  This is a first try to learn what will go wrong.Steppers came from: MIT flea market in Cambridge MA, another surplus store in Cambridge which I think is gone and cannot recall the name, ELI surplus in Framingham MA (gone) surplus places in Canada, and free samples.This has been going on for some time obviously.  For controllers the hard choice is using stuff that is stacked all around, or buying something new. No idea how that will turn out. Yes luck will be needed, thank you.Best to use common sense before doing anything shown above.
Reply:What did you end up using for controllers? I have a couple of Slo-syn motors and Slo-Syn 3180-pt but I am not sure how to hook it all up or what to do with it.
Reply:Originally Posted by gow589What did you end up using for controllers? I have a couple of Slo-syn motors and Slo-Syn 3180-pt but I am not sure how to hook it all up or what to do with it.
Reply:4 wire. The motors are:M093-ff-4024.5v 3aI understand with most stepper setups there needs to be a controller board.....but I also read somewhere that the 3180 needs to be daisy chained so I am really unsure how it all goes together. I do dable in electronics but I have never worked with steppers.
Reply:Subscribed. Waiting for the finished product.Good Luck.We are on our way to being farmers! Our site is always under construction so check back often for updates.  http://www.philosophyfarmstyle.com/
Reply:Originally Posted by gow5894 wire. The motors are:M093-ff-4024.5v 3aI understand with most stepper setups there needs to be a controller board.....but I also read somewhere that the 3180 needs to be daisy chained so I am really unsure how it all goes together. I do dable in electronics but I have never worked with steppers.
Reply:socoj2, that is interesting... but my new 420 lenovo thinkpad does not have p-port OR pcmcia card slot.... my old Toshiba does have the card slot.... is there some other way to control it ?Weldandpower Lincoln 225 AC,DC with Briggs 16hp gas engine.WW2 era Miller TIG.
Reply:When I read the specs on that card I did not see the pcmcia requirement. It wants a parallel printer port and most PCs, even laptops, still have these, or you can buy a usb - printer port external module to get one.CharleyMiller MM252Miller Bobcat 225NTMiller DialArc HF / DIY Cooler2 Victor O/A TorchsetsMilwaukee 8" Metal SawMilwaukee Dry Cut "Chop" Saw 5 Ton Wallace Gantry Various Grinders, Benders, etc.
Reply:new computers dont have printer ports, i have a computer i built 3 yrs ago that doesnt have a printer port, i used a typical manufactured motherboard..  you can add a port card.. there is also a usb driver board setup that is out but its a bit more expensive... you want an older computer probably around 2ghz processor and atleast a 1 gig of memory... i'm using an allen bradley 2.2ghz pentium 4 with 1.5gig of pc2700 memory, onboard audio, onboard video, on boardnetwork card. i built it from a tub of scrap computers, it took me 2 computers and various components to make 1 decent computer.. i have the computer set up with a dual boot setup.. one operating system is ubuntu 10.04 with emc2 included (linux cnc control software), I also have windows xp installed with a windows based cnc control software... i will run an external hard drive for storage... i started on my table yesterday, i have the top part of the perimeter sawed, milled to length, drilled & counterbored&tapped... i also made some corner brackets and bolted it together... not much but its a start.. right now with the computer built with cnc control software, and the perimeter of the frame built i have $0 invested in it.. Attached ImagesLast edited by brucer; 09-17-2011 at 12:10 AM.tackleexperts.comwww.necessityjigs.comhttps://www.facebook.com/groups/mach...dingequipment/
Reply:dont want to hijack the thread, but Brucer can you tell me more and where you will be getting parts from, I use to go to another site but cannot remember 1 of the machinest sites and had a lot of good info on home built.  I am thinking about building one and want to gather all the info, again sorry for butting in.ron
Reply:cnczone is probably the site your looking for..... i own my own mill so i'm making all the brackets and skates and rails myself, i will probably run a small chain drive as its readily available locally.. i will probably get the chain,sprockets,pillow blocks and clamp collars from my local Rural King agriculture dept...I'm thinking like 3/8" or 1/2" chain drive.. i know chain drive isnt the most accurate drive system, but its readily available locally, and from pricing i think its the cheapest with belt drives coming in just a little bit more expensive.. it would be easy to convert to belt drive later on also.. i think the chain drive will work well enough for me here at home..if i were running a business i would probably go with either acme leadscrew or rack drive, more than likely rack drive..Last edited by brucer; 09-17-2011 at 03:05 AM.tackleexperts.comwww.necessityjigs.comhttps://www.facebook.com/groups/mach...dingequipment/
Reply:Thanks I went searching and found the site, thanks again
Reply:Some times CNC zone has so many sub forums post just get lost. It is good for research though.  Originally Posted by brucercnczone is probably the site your looking for..... i own my own mill so i'm making all the brackets and skates and rails myself, i will probably run a small chain drive as its readily available locally.. i will probably get the chain,sprockets,pillow blocks and clamp collars from my local Rural King agriculture dept...I'm thinking like 3/8" or 1/2" chain drive.. i know chain drive isnt the most accurate drive system, but its readily available locally, and from pricing i think its the cheapest with belt drives coming in just a little bit more expensive.. it would be easy to convert to belt drive later on also.. i think the chain drive will work well enough for me here at home..if i were running a business i would probably go with either acme leadscrew or rack drive, more than likely rack drive..
Reply:Brucer, Thanks for those specifics on the computer and control software.  I had added a PCMCIA card to my desktop once I had to invest in stuff which fit it for my old laptop.   When you mentioned the Acme thread possibility in the future... you meant direct (end) drive  of the Acme rod for accuracy ?     With some CNC procedures speed would not be a problem with regards to quality... like a wood router movement.... but with the plasma cutters there is some need I think to hit a basic control speed of travel  for any particular thickness and type of metal for the torch head being used..... So the question is... using Acme threads...which for a straight line cut  in 1/16 th inch clean steel.... would be moving pretty fast in terms of the rpm's of the Acme threaded rod... are there plenty of reasonable options available for getting those kinds of speeds ?  Are we talking , in the initial design process decisions , more voltage available , bigger faster steppers in the first place ... or what ?   I would like to avoid chains and belts myself.... and gears have drawbacks also....    Are there charts made up to look at to plug in some of these types of needs which show the legit options for meeting them in a proper machine design ?Weldandpower Lincoln 225 AC,DC with Briggs 16hp gas engine.WW2 era Miller TIG.
Reply:Originally Posted by GBMBrucer, Thanks for those specifics on the computer and control software.  I had added a PCMCIA card to my desktop once I had to invest in stuff which fit it for my old laptop.   When you mentioned the Acme thread possibility in the future... you meant direct (end) drive  of the Acme rod for accuracy ?     With some CNC procedures speed would not be a problem with regards to quality... like a wood router movement.... but with the plasma cutters there is some need I think to hit a basic control speed of travel  for any particular thickness and type of metal for the torch head being used..... So the question is... using Acme threads...which for a straight line cut  in 1/16 th inch clean steel.... would be moving pretty fast in terms of the rpm's of the Acme threaded rod... are there plenty of reasonable options available for getting those kinds of speeds ?  Are we talking , in the initial design process decisions , more voltage available , bigger faster steppers in the first place ... or what ?   I would like to avoid chains and belts myself.... and gears have drawbacks also....    Are there charts made up to look at to plug in some of these types of needs which show the legit options for meeting them in a proper machine design ?
Reply:Thanks Brucer.Very interesting and gives me lots of thing to research..If you want to take out the slack in your chain drive... you can place two sprockets on the motor AND the driven rod.... ... and use idlers placed on opposite sides ... pretty much a no slack situation... spring load those idlers.... That gives you the option if you need to of adjusting the relative speeds later by only changing sprockets... but is pretty much a no backlash arrangement...Weldandpower Lincoln 225 AC,DC with Briggs 16hp gas engine.WW2 era Miller TIG.
Reply:Originally Posted by CharleyLIt wants a parallel printer port and most PCs, even laptops, still have these, or you can buy a usb - printer port external module to get one.
Reply:Originally Posted by TensaitekiNo. If a driver card uses the parallel port, it likely sends step and direction signals and a USB - parallel adapter WILL NOT WORK. Real-time step/dir signals cannot be sent through the usb port.[/LIST]
Reply:heres a nice belt driven 5x4 table  of course hes running some inexpensive linear rails he found at a salvage yard, wish i could find some as cheap...   it is a nice build though.. he did a nice well thought out job.. I think he said he was running 305oz steppers..  hes focused on routing, but it would be esy to add a water table..  as Tensaiteki said, some of his reasons is why i kept my table on the small side (50x40), plus i dont have room for a large table,  but i am planning on using chain drive, if it works its cheap, converting to rack would be pretty easy later on so i'm not that worried about it.. heres another clean looking design, i think its rack drive but with the cables it might be cable driven. hard to tell on my laptop. heres a hybrid belt leadscrew design. Last edited by brucer; 09-18-2011 at 01:14 AM.tackleexperts.comwww.necessityjigs.comhttps://www.facebook.com/groups/mach...dingequipment/
Reply:Originally Posted by brucerthey have usb driver boards now, i've seen a few places selling them.  http://www.planet-cnc.com/
Reply:Tensaiteki ,  GOOD STUFF !!!!!!   Thanks for taking the time to share that.Weldandpower Lincoln 225 AC,DC with Briggs 16hp gas engine.WW2 era Miller TIG.
Reply:Originally Posted by TensaitekiThat is not a driver board, that is a controller board, there is a difference. That USB controller still needs separate drivers to turn the motors.When using a parallel port based system, the PC is the controller, it reads the G-code (interpreter), plans out the accelerations and movements (trajectory planner), and sends the step/direction signals to the driver at the appropriate time (step generation). The reason the parallel port is used is because it allows you to send different signals on different physical pins simultaneously. These step and direction signals must reach the drivers in real-time, uninterrupted and without delay, with micro-second (not simply millisecond) precision. The USB port is inherently incapable of real-time communication, thus USB-parallel adapter cables will not work for drivers that expect step and direction signals.Hardware controllers are not new. The USB controller you posted, is just that, a controller. It is simply fed the raw g-code through the USB interface from the PC, and handles the interpreting, trajectory planning, and step generation in-hardware. Aside from drip-feeding G-code, all the PC is for is to provide a user-interface.The biggest problem with these hardware/micro-controller based controllers is that they are not as powerful as a software based controller running on a full PC. This is especially important for plasma systems which require advanced constant-velocity trajectories on parts that are broken up into hundreds or thousands of small segments. To do this the controller must be able to look-ahead hundreds or thousands of lines of G-code and preform serious mathematical calculations while controlling the motors at the same time, this is easy for a full PC, not so for a measly little micro-controller. They are also far less customizable for applications they were not specifically designed for (such as plasma).Controllers for plasma machines must be specially set up to handle the real-time feedback from the THC. It does not appear that that board is capable of accepting those signals and  it is extremely unlikely that it is capable of using them properly if it did. A properly integrated plasma controller will not simply move the torch up or down to maintain a set voltage/height, it will also monitor the actual speed of cutting and lock the torch height if it slows down too much to keep the torch from diving into the plate on corners. It will also monitor the arc voltage for rapid changes and lock the height in the event of a voltage spike (usually caused by cutting across a gap or previous cut) preventing a torch dive. These functions are relatively easy to implement in a software controller (they already exist in EMC2 and Mach3) but far more difficult in hardware controllers.Sure you could add a water table, but it still wouldn't be a very good plasma machine. It's easy to tell from the sound and motion in the video that he has very low acceleration, high speed, but low acceleration. (In the CNCZone thread he even says that the acceleration is low). Low acceleration is fine for routers, but not for plasma machines. He is driving that big gantry with one little motor, no wonder his acceleration sucks.The HobbyCNC kit he is using is limited to 36V and 3A per motor. Higher current motors give more torque (especially at low speeds) and thus better acceleration. Higher voltages give higher top-end speeds. This is why I say not to d!ck around with small motors, small motors can give high speeds, but only at cost of acceleration, or high acceleration but at cost of top speeds. Plasma needs both not either/or.It is rack drive. The cables are there too keep the gantry "square" because he is only driving one side of the gantry. Again, he is using small motors (one motor driving the gantry) and has very low acceleration.If you watch the video of that machine cutting, it is obvious that he is running with the constant-velocity (CV) mode turned off. If he didn't, his low acceleration would cause the controller to severely round off all sharp corners in an attempt to keep the speed up. However, turning the CV off will produce a divot in the cut everywhere the machine had to decelerate, stop, and accelerate at the end of each cut segment. It also results in the wildly varying cut speed evident in that video.That's not hybrid, both X and Y-axis are leadscrew driven. Hybrid would be if one axis was leadscrew driven and another was pure belt driven, simply driving a leadscrew with a belt does not make it a hybrid.That machine is also a good illustration of why leadscrews are not very good for plasma machines. As you can see, the leadscrew provides excellent acceleration but poor top speeds. Again, if you watch the video of the machine actually cutting, you can see that they are cutting what appears to be thin aluminum sheet but way too slowly. Cutting too slowly results in excessive dross, poor cut quality, excessive heat input, wider kerf (less detail possible), and more warpage.
Reply:Originally Posted by TensaitekiThat is not a driver board, that is a controller board, there is a difference. That USB controller still needs separate drivers to turn the motors.When using a parallel port based system, the PC is the controller, it reads the G-code (interpreter), plans out the accelerations and movements (trajectory planner), and sends the step/direction signals to the driver at the appropriate time (step generation). The reason the parallel port is used is because it allows you to send different signals on different physical pins simultaneously. These step and direction signals must reach the drivers in real-time, uninterrupted and without delay, with micro-second (not simply millisecond) precision. The USB port is inherently incapable of real-time communication, thus USB-parallel adapter cables will not work for drivers that expect step and direction signals.Hardware controllers are not new. The USB controller you posted, is just that, a controller. It is simply fed the raw g-code through the USB interface from the PC, and handles the interpreting, trajectory planning, and step generation in-hardware. Aside from drip-feeding G-code, all the PC is for is to provide a user-interface.The biggest problem with these hardware/micro-controller based controllers is that they are not as powerful as a software based controller running on a full PC. This is especially important for plasma systems which require advanced constant-velocity trajectories on parts that are broken up into hundreds or thousands of small segments. To do this the controller must be able to look-ahead hundreds or thousands of lines of G-code and preform serious mathematical calculations while controlling the motors at the same time, this is easy for a full PC, not so for a measly little micro-controller. They are also far less customizable for applications they were not specifically designed for (such as plasma).Controllers for plasma machines must be specially set up to handle the real-time feedback from the THC. It does not appear that that board is capable of accepting those signals and  it is extremely unlikely that it is capable of using them properly if it did. A properly integrated plasma controller will not simply move the torch up or down to maintain a set voltage/height, it will also monitor the actual speed of cutting and lock the torch height if it slows down too much to keep the torch from diving into the plate on corners. It will also monitor the arc voltage for rapid changes and lock the height in the event of a voltage spike (usually caused by cutting across a gap or previous cut) preventing a torch dive. These functions are relatively easy to implement in a software controller (they already exist in EMC2 and Mach3) but far more difficult in hardware controllers.i seee what ur sayin with ur build brucer.   ur building a budget plasma table  and will have plenty of room for upgrades.     and we all like improving things. .   and adding new parts.     i think what ur doing is what i would likely build.    and its how i would learn and gain experience with cnc control and programing.   u definitly seem like  u know ur stuff as well.>Innovations are what i leave behind for History
Reply:here is a large chain drive mechmate..  huge compared to the machine i plan on building, but another nice clean chain drive machine nonetheless.  here is another chain drive machine, i dont care for the design of it though, and i dont care for wood machines, one of his steppers sounds flaky to me.. but it seems to run good and looks pretty fast to me.. again its a bit larger than what i'm building....Last edited by brucer; 09-19-2011 at 02:04 PM.tackleexperts.comwww.necessityjigs.comhttps://www.facebook.com/groups/mach...dingequipment/