[banjopup]

Here is what I would like to do. I currently have a cyber-nc router which works pretty well. Because I am a cheapskate I would like to see if I can use the controller and motors (or at least the controller) to build and run a 5x10 plasma table. I have an ESAB Powercut 875 that I would be using on this machine. Any thoughts, advice, plans etc?

[Torchhead]

Here is what I would like to do. I currently have a cyber-nc router which works pretty well. Because I am a cheapskate I would like to see if I can use the controller and motors (or at least the controller) to build and run a 5x10 plasma table. I have an ESAB Powercut 875 that I would be using on this machine. Any thoughts, advice, plans etc?

Is it a 4 axis controller? Will it run with a PC and MACH3? With a 5 ft cutting area (6 + ft gantry) you need two things for plasma: speed and acceleration. Those call for a lighter gantry (acceleration = force/mass). Every pound you take out of the equation buys you more acceleration. Every IPS?sec more acceleration you have the sharper the turns your machine can do. So big heavy gantries are great for routing and milling but need big heavy motors for plasma.

The downside of a light gantry is you get flex over the length. If you drive a 6 ft long lever from one end and do the acceleration you need in plasma then you can see inertial forces at work with some ugly cuts. The solution is to drive both ends of the ganty. You do that either mechanically with a cross shaft and one big motor or twin motors "slaved".

The MACH question has to do with the inevitable question about what you do about Torch Height on that large a cut surface. It's disappointing to find you saved money on the build only to find you can't take advantage of the cost effective THC options.


TOMcaudle
www.Candcnc.com

[banjopup]

Tom,

Sorry I haven't responded sooner. Thanks for the reply BTW. After doing some more reading on the site I have decided not to go the 'cheapskate route' (especially since my little router is doing a little paying work right now) but rather to build a new one.

I have Mach3 and BobCad (both of which I am still learning to use).

The build that dnelso has going on in the post about the two tables he is building (or did build) is the one I'd like to model mine after. I still need to pinch pennies everywhere I can of course.

Thanks,
Steve

[ratfink]

Hi Banjopup,
I am in a similar situation. However I need to build a Plasma Table that is 5x20 feet. We are in the same boat.
My concerns are I have no issue using big heavy motors... should we use Servo or Steppers?

Another possibly silly question would be this.... Back in the old days when I was messing around with robotics... I was using Electric Wheel Chair Machine Motors with seperate encoders attached.... These motors are extremely powerful, they are DC, so variable voltages are common for speed control. The encoders are cheap and very accurate. I have not found anyone who has gone this method and I am guessing there is a reason?

Next Question: I see everyone driving off of a belt with reduction gearing? seems like a lot of extra work with brackets, sprockets, etc...., It would seem to me that these huge tables with these large motors should take advantage of direct drive with Chain, and or Screws. Or direct drive tracks at the very least? I understand the need for gearing in order to take advantage of motor sizes, torque etc... however no one explains what thier doing or the calculations they are making with sproket sizes etc.... its like everyones doing a hail mary pass at this point in all thier build to figure out what works best. It does no good for anyone on the forum looking at redesigned, brackets that are all cut up multiple times,,,, different sprockets, changes,,, etc.... and sending more confusion our way.... It makes more sense to spend a little more money on larger motors and direct drive or as close to a known drive geared formula with a simple parts bracket listing than have to design the taj mahal multiple times.

I get it... I understand speed, torque, vibration, smoothness (For lack of better words) weight, height, of the business end of the plasma machine... Im also surprised that everyone is using this strange square linear actuator plaform with four small bearings on each end that ride on a square 2x4 piece of steel?, there is no way that is even close to holding any type of tolerance with the gantry. I have seen this setup and unless it is part of an entire assembly (Meaning track, bearings, and platform that bolts to the top of the 2x4 steel tubing) it will never hold close tolerances. Personally I have used circular slotted steel wheels on 2" round tubing with accuracy to .003, and capable of holding very heavy gantry assembly at very inexpensive prices... the highest priced part is the wheels... essentially conveyor belt wheels which last forever.

Seems to be it would be simpler to use big inexpensive DC servo motors, slapp encoders on them, with a simple direct drive, and track system that will hold the weight, strength , rigidity of the system.

My appologies in advance if Im a complete idiot... You can source a used electric wheel chair for a hundred bucks. encoders,,, 30 each, you can use the wheel chair control stick in place of a joystick and or wire it the same, dont have to deal with PS, as you will be using all the parts in the wheel chair.. however, the wiring of the encoders, etc... will have to be figured out... but all this has to be figured out anyway, right? Just seems like we are missing some other really inexpensive options out there... Please feel free to trash my ideas and beat me up.... I would really like to be educated and certainly do not intend to offend anyone..... Maybe I havent looked or read enough in the forums? But again,, I appreciate any help or ideas to these suggstions..... for all our sakes?

[Torchhead]

Usually if no one has done it before in a market that is 8 years old (DIY/Hobby/SmallShop) it may be a clue. There are a multitude of issues with controlling servos not all of which are just torque and speed. Also the challenge of building a complex CNC machine or of "picker" parts may have an appeal if your time is cheap. A lot of builders at more into the destination (getting a plasma table up and making parts) than the journey and learning a new set of skills (electronics tech, programmer, design engineer, etc)

i would say that if you are driven to prove it can be done than by all means push on. You may someday be hailed as pioneer.

One thing you need to try and bring into your thought process: Non-contact flame cutting does not benefit from bigger, heavier, massive components. I always use the "Sports Car VS Dump Truck" argument. each has a different role and job to do and building a Sports Dump Truck will be a compromise and more costly in the long run.

The gearing of a machine is always a dance between speed and torque/resolution. While servos have a better dynamic range.so function better where you need to do mixed types of cutting (contact & non-contact) the equation still exists. If you give up torque and resolution for speed (even if you don't use it) you cannot recover that side of the equation. So sit down with a calculator and run some numbers.

Take the shaft RPM (max) of the motor and multiply it times (PI X Dia) of the FINAL drive element. If that is a sprocket, gear or whatever. Then you have you speed setup ratio and the resolution step down ratio. You loose linear force (that moves and accelerates the gantry) by the radius of the final drive

Servo motors are just DC motors with an encoder although the brushes are set differently so it is not an exact comparison. Slow speed operation of a low voltage DC motor takes lots of amps and the drives have to handle that. batteries are real good and providing lots of current at low voltage but also with limited run times.

First thing you have to do is find motor drivers that will take conventional control signals (i.e. Step & Dir or + & - analog) and control the motors. Since smaller servos operating at higher voltages ( remember: Voltage times current = watts = horsepower = work) draw as much as 20 to 30Amps during acceleration or if under max load) You may need drives that could handle 60, 80 or 100A without blowing.

Don't take this as a "Don't even try it" answer. After all early rocket guys were laughed at. Just realize you can't cheat Physics so get a good understanding of the forces and what the final motion really needs to do.

TOM Caudle
www.CandCNC.com

[ratfink]

Hey Tom,
Yes, makes perfect sense. I dont want to accidentally highjack this thread so I am in hopes this info will be helpful if not at least a good review of basic principles.
Thank you for the detailed info, its very helpful.
it now looks as though I will be building a plasma cutter that is approximately 12'x40', I need to see if anyone has plans out there for a commercial size as this.
Thanks again
J

[Torchhead]

Hey Tom,

it now looks as though I will be building a plasma cutter that is approximately 12'x40', I need to see if anyone has plans out there for a commercial size as this.

that's not a plasma table, that's a plasma PLACE. I've had shops smaller than that!

Still, the issue remains of force and mass. The total weight of the gantry is the mass you are moving (accelerating from 0 to full feedrate) length has no bearing on the physics side. You do have to rethink the electronics side operating over lengths exceed 55 ft. Conventional encoders even in shielded cable will be useless in a plasma noise environment. High current motor wires with most drivers (PWM modulation) become 100W+ LF transmitters.

Not sure I would recommend that as a DIY project especially if it is your first plasma CNC build.

Nothing is a good deal if you can't make it work.

TOMcaudle
www.CandCNC.com