Robot Project (Need some help-please)

[nihaddad]

 ;DI want to create a robot arm that solders wires onto pins for an application that I use at work. However, I am doing this on my free time. But I don't know where to begin. I read the tutorials and bought books on electronics and robot building.

Application:

Solder very small fixed wires onto pins (SEE attached PICTURE)

Should I focus on the programing first, electronics, or mechanical aspects? I have some draft ideas, but am worried that my drafts are too complicated. My robot arm needs to do many things hold a wire in a location, hold solder wire, and operate a soldering iron.

Basic design ideas:

The soldering tip will be the end effector with 2 DOF
The soldering wire will be controlled at the end effector by some type of mechanism that continuously feeds wire in a specific location (probably add a dowel to hold wiring that can be feed into a machine that rolls the wire down a track through an applicator wand)

But what about my wires that I need to solder on. Would I need to make another robot arm or is there a way to integrate the wire on the effector?

More Questions:

How accurate can a robot be at sensing wires that are about 30 gauge  or smaller and then grasping said wire? Can it be done?
Is this project at all feasible?
Where do I even start?

I am well aware that this is a very ambitious project, that might cost over $2000 and take a year to complete.

And did I mention I have never built a robot before, have no idea how to program, and have only a rudimentary understanding of electronics.

[Anoroc]

HI there,

I'm no authority around these here parts, (just started my 1st robot), But so far my experience has taught me to keep it simple! I can only recommend you think about starting with the $50 robot project and work from there. From what I've read, a robotic arm is a little overly-ambitious for a first-timer.

[amando96]

HI there,

I'm no authority around these here parts, (just started my 1st robot), But so far my experience has taught me to keep it simple! I can only recommend you think about starting with the $50 robot project and work from there. From what I've read, a robotic arm is a little overly-ambitious for a first-timer.

A robot arm is pretty simple to be honest, but one with that requires all that precision is much more difficult, and would require very expensive servos.

[nihaddad]

I have a pretty large budget and am extremely motivated. But what servos would I need? And do you think i would have to build two arms to handle the wiring, solder wire, and solder tip?

[dellagd]

My opinion, Just have a electronically adjustable clamp to put your board into, then just move the board around a fixed solder tip! much more simple.

Just have the board be X and Y axis adjustable, then, whenever going to a new joint, have it drop down, adjust position, then rise back up into soldering tip. Much more simple.

[TK]

I have built three different robotic arms in the past.  One using servos to move a tennis ball, another with motors for picking up pvc pipes and the last is with stepper motors and it is supposed to replicate a robotic arm used in a factory.  I find that the stepper motors are the trickiest to get used to but once you have it down pat they are the most precise out of the three.  With three steppers you could easily build an arm or even make your own  version of a CNC machine

[Soeren]

Hi,

I have a pretty large budget and am extremely motivated. But what servos would I need? And do you think i would have to build two arms to handle the wiring, solder wire, and solder tip?

You should start with planning, making sketches of different ideas and then put them aside for a week or two (while doing other parts of the project). The subconscious is a great problem cruncher and when you go back to the sketches, you will have reached some conclusions, seemingly without any work.
Then make simple prototypes of the best ideas, but keep it simple, you shouldn't build the entire thing, just sub-assemblies, to make sure that it will work as intended - sometimes a cardboard prototype will do.

Like your mention of a 2 DOF arm to control the soldering iron...
If you made even the simplest model (in cardboard, Mechano, LEGO or whatever), you would realize that 2 DOF isn't really gonna do it - you need to be able to move it in 3 planes to position it (X and Y) and lower it to the joint (Z).

And did I mention I have never built a robot before, have no idea how to program, and have only a rudimentary understanding of electronics.

Now this statement makes the rate of success somewhere between slim and none and my best advice i to buy either a finished robot meant for this app. or hire a skilled engineer to do it.

Had you written something like: "I need it to solder a wire with a diameter of 0.02 mm (tol. 0.001 mm) to a pad measuring 1 mm +/-0.03 mm..." etc. I would have determined your chance of success a fair bit higher, as this is how engineers approach a problem.


So, for starters, throw some numbers around and think in modules: One "arm" to feed the wire, on arm to place the soldering iron, a third "arm" to apply solder and when each is doing a proper job, you can build them into a combined robot.

For detecting the position of really thin wire, a linear opto-array is the way to go - they're used to detect irregularities in thin tread in industrial weaving machines among others and they can be had in resolutions in the few µm range.
The same array could be used for positioning it precisely over the part, used as feedback in the servo circuit you're gonna need (don't confuse this with R/C servos, they wouldn't be anywhere near the precision you need).

You need something to aim the soldering iron as well. A LASER pointing to the pad could be used for easier aiming, if you want to move the iron around, but as someone mentioned, an X-Y plane moving the board around under a soldering iron that just have to be lowered.
You also need to figure  way to clean the soldering iron in between each soldering.

The best "robot" for such tasks is still young girls (probably since they'll work at way below minimum wages in some contries, just to keep hunger at bay) - even the dumbest person will be way better than anything you can build in a year with such a modest budget - $200,000 is a more realistic budget for such a project, at least if you want a (western) company to make it.


Building robots, you need to be skilled in mechanical and electronic design, as well as programming - remove one and the project won't be done the way you intended it.

Just took a second glance at your photo and it still doesn't tell me what it is that you want the robot to solder?

[nihaddad]

appreciate the feedback-i definitely will get all the measurements worked out before I begin planning.