I have been going through the theory of constructing and programming a robotic arm based off of either flex sensors or EMG sensors. I have looked around and seen that string controlled by servos is prevalent in these kinds of robotic arms. The thing that I am struggling to figure out is how I will be able to have the arm actually lift weight. I have been looking at different servos, but I don't think that they are strong enough to lift, say, a 15 pound weight very easily. I know that this is quite a bit, and have been looking at other alternatives as well. I have seen Easton La Chapelle's video of his robotic arm and he used a DC motor for the shoulder and elbow joints. The thing with this is I don't really understand how the feedback with the potentiometers and how to control the position of the motor itself. I want it to be able to change position based on the flexing/bending of the arm (which is where the EMG and flex sensors come in) and thus have the DC motor essentially act as a servo. I'm hoping that there is a DC motor that provides enough torque to be able to lift 15 pounds (more than this may not be necessary), and what I desire to know are the components needed to be able to control the position of the motor to do so.
I have looked around and seen that string controlled by servos is prevalent in these kinds of robotic arms.
Do you have a picture or video of the type of arm you are talking about? You might use linear actuators like the below digger to make a large arm. If you have some $$$ for the project you could modify the bottom trencher. 15 lb is a lot of weight in the hobby robotic arm world.
These servos are cheap and powerful
http://www.servocity.com/html/spg785a-5_0_servo_gearbox.html
Attached is is picture with the scale of lift force of the servo.
The downside to them is they are bulky and require shielding around the aluminum gear to prevent damage. Fully geared down they are able to life a 6 cans of cola at 18" If you linkage them into a 4 bar mechanism you could easily double that.
On the Sketch Up 3d warehouse I loaded a 3d file to help with servo design in mechanisms. It is found under the name "Servo Tool kit"
It shows many different servos, their cost, their torque (lifting cans of cola at a certain arm length), and their type.
http://sketchup.google.com/3dwarehouse/details?mid=bd8897b1471e2d1c73ac8f193878e4bc&prevstart=0
-pb
zoomkat:
I have looked around and seen that string controlled by servos is prevalent in these kinds of robotic arms.
Do you have a picture or video of the type of arm you are talking about? You might use linear actuators like the below digger to make a large arm. If you have some $$$ for the project you could modify the bottom trencher. 15 lb is a lot of weight in the hobby robotic arm world.
FIRGELLI | 12v & 24v Electric Linear Actuators
http://www.harborfreight.com/towable-ride-on-trencher-65162.html
Indeed, 15 lb is, and would probably be leaning more into more industrially-prevalent methods. I have considered the use of linear actuators and am only concerned with the design, as a motor would seem more convenient in that it is closer to the movement of the joints in rotation. The arm that I want to eventually construct will ideally mimic the movements of the human arm, so using actuators alone may not be the most preferred. Perhaps a linear actuator would work in combination with a motor to provide the rotation while also providing the lifting force... though I admit I'm a bit unqualified to state how exactly the mechanisms would work practically.
piratebrian:
These servos are cheap and powerfulhttp://www.servocity.com/html/spg785a-5_0_servo_gearbox.html
Attached is is picture with the scale of lift force of the servo.
The downside to them is they are bulky and require shielding around the aluminum gear to prevent damage. Fully geared down they are able to life a 6 cans of cola at 18" If you linkage them into a 4 bar mechanism you could easily double that.On the Sketch Up 3d warehouse I loaded a 3d file to help with servo design in mechanisms. It is found under the name "Servo Tool kit"
It shows many different servos, their cost, their torque (lifting cans of cola at a certain arm length), and their type.
http://sketchup.google.com/3dwarehouse/details?mid=bd8897b1471e2d1c73ac8f193878e4bc&prevstart=0-pb
I actually have never seen those servos before... and the maximum torque kind of blew me away, as I haven't actually delved too much in gearboxes. In this sort of project, power actually holds a priority over profile, as I know some sacrifices must be made in the exchange of size for power. I am thinking if I were to use these, I would definitely have to research some on the shielding that you mentioned just so I don't screw up such a product, as I didn't actually know about the requirements of such protection. I appreciate the information, as well as the model you have provided, and will definitely be taking a look at it once I download Sketch Up. 
As a student right now, cost is definitely a factor in purchasing the components for the project. However, if I do become successfully able to finish the functioning project, I think spending a few hundred dollars would be well worth it in the long run. I really do appreciate you guys helping me as I'm actually hoping to create a sort of assistant for those who may be unable to perform certain every-day tasks that may involve some weightier items. I essentially picked 15 pounds as an example weight because I was thinking of a watermelon, though I am fairly sure it would be difficult for a robot arm to both lift the watermelon with one hand and keep it balanced at the same time... but of course, such an assisting machine, were it to be successfully built, could be programmed to complete other tasks too. If anyone has any alternatives that may actually be better suited for this kind of project, feel free to let me know, as learning throughout this process is just one of the many benefits I gain from attempting to do this.
(I also can't wait to use a bunch of batteries to power everything as well. :1 )
Search on youtube for things like robotic arm and servo arm to see what others have done.
Those particular servos are able to take direct signal from the arduino (but use a separate +6v power supply) because they are simple winch servos with +5 full rotations. Every angle is then programmable with the simple writemicroseconds() command.
Any servo that handles even a marginally useful amount of force (two text book worth of weight at at 1foot arm) is going to be $100-200 each.
The linear actuators you mention (source below) are not amazing in their speed.
http://www.robotshop.com/firgelli-actuators.html
Pneumatics may be the way to go but with any pneuamatic project the cost quickly gets high due to part specific purchases guiding the troubleshooting.
You can scrap together a large working arm with hospital bed motors, they are very easy to get, but AC.
Good luck!
pb