Conceptual idea

bfgstew

bfgstew

Senior Member
Maybe not new but this has been rattling around my head for a while and was wondering if it is at all possible?
While designing and building my CNC machine (still in the mechanical stage of construction) and watching two robots (at my work place), the thought/idea came to me about building my own robot arm, now, nothing new in that but I thought a bit longer and how about if a robot arm could do the same job as a CNC machine?
Using the various steppers I have aquired (4.6A to 2.6A) units of high torque, one for rotation of the base coupled to one for the first arm another at the end of that arm to move the next arm and another on the end to move the 'hand' and maybe another on the hand to 'hold' various tools. As the title of this thread states 'conceptual' it may get shot down in flames or may start a decent debate on the merits of powerful robot arm usage and limitations with the PICAXE or PICAXES as the brains. Mechanically it will need to be rigid enough not to flex or bend but light enough for the steppers to hold with its own torque. Watching our robots at work is fascinating, the accuracy and speed is remarkable to say the least.
I will start to draw up some rough sketches and access material availability and strength, to see if it is possible, well I know building a robot arm is possible so if the CNC route is a failure at least a robot arm will come out of it.

Look forward to hearing your views.

Stewart
 
hippy

hippy

Ex-Staff (retired)
The idea seems sound. Moving a CNC table and keeping the tool locked in position, or keeping the table locked and moving the tool, achieves the same thing. You could even move the table along only one axis, move the tool along the other.

The decision probably comes down to the advantages and disadvantages of each. Possibly that accurately moving a table is easier for a light part than moving the heavier tool head.
 
Last edited:
C

Captain Haddock

Senior Member
I guess the advantage of moving a bed over moving an arm with tool is not having to deal with torque reactions from the spinning tool.
Will be interested to see how it pans out.
 
S

StigOfTheDump

Senior Member
I think you will struggle to get accuracy particularly with the base rotation. A one degree step would translate to 9mm of movement at a 500mm radius. Gearing would not help as all backlash would be multiplied at the tool. One way would be to drive a screwed thread through the end of a radius arm, but the amount of movement at the tool would vary according to the angle, and it wouldn't be able to go 360'. Epicyclic gearing would give the best reduction with the least backlash.

Also a major challenge would be the programming. Starting with linear dimensions, then converting to rotary movement to end up with a linear movement.

If the arm was mounted from above it would be brilliant for doing circular work and spirals, but not so hot on straight lines.

If it was only used to perform an action after relocating itself, such as drilling a hole then I think it may be useful. For milling etc I think it will be less so. A rotary arm on the toolhead of a traditional CNC table might be a better option.

I will be interested to see how you get on.
 
Last edited:
bfgstew

bfgstew

Senior Member
StigOfTheDump said:
I woke up this morning thinking worm and wheel.
Click to expand...
You need a holiday..............LOL

Appreciate early feedback, and there are going to be a lot of issues that would need to be addressed before any point in starting and the biggest is programming the thing, yes fairly easy to get an arm to 'do a set routine' but to get it to read and convert say an outlay from DesignSpark to mill a board is going to challenging to say the least.
Backlash shouldn't be an issue with steppers and the radial distances can be eliminated by having an extra stepper on the end of the first arm to articulate it through 90 degrees opposite to the stroke arm as we have on our ones at work.
As i say, early days, lots to think about, lots to solve.
 
marzan

marzan

Senior Member
You would have a lot of vectoring to do as well. Moving from, say. x3y3 to x3y20 would be a straight move for a CNC mill. But if you did id t with a robot arm it`s not that easy. I thought about doing the exact same thing, but ended up converting a drill mill and then using that to build a gantry mill.
Marz.
 
hippy

hippy

Ex-Staff (retired)
marzan said:
You would have a lot of vectoring to do as well. Moving from, say. x3y3 to x3y20 would be a straight move for a CNC mill. But if you did id t with a robot arm it`s not that easy.
Click to expand...
Getting a robot arm to draw straight lines, circles, triangles and squares is probably a good start to solving the maths problems.
 
You must log in or register to reply here.
Top