# Effective motor type for robotic arm?

I am trying to build an arm that will have about 5 by 5 by maybe 7 or so centimeters of room for a rotary motor capable of lifting it. The joint will basically allow the arm to rotate in a single degree of freedom in a circle that is perpendicular to the ground.The rest of the arm will probably be around 64 centimeters long and weigh around a minimum of 9 kilograms before it lifts anything.

What kind of motor type would give it the best chance of lifting the arm quickly and reasonably accurately?

† Raising from straight down to out 90 degrees in around 1 to .5 seconds maximum.
‡ At least a centimeter at the end of the arm which means probably at the very least 300 positions for the motor.

• Please edit question to answer following. Does “5 by 5 centimeters of room for a motor” mean the motor face can be 5x5, and motor length is unconstrained? How many degrees of freedom? Is motion rotary, hinge, or what? Eg does arm rotate on a base, or move more like an elbow? Sep 10, 2013 at 20:51
• @jwpat7 That is a very good point, I will add that now Sep 10, 2013 at 23:07
• I don't think there is enough information in this question for an objective answer, to the point where it might have to be put on hold. How fast does the arm need to move in order to meet your needs? How accurate does it need to be? From that, you can calculate the needs of your motor -- and from that, you can see whether you have any options in the size you are hoping to find.
– Ian
Sep 12, 2013 at 17:00
• What Ian says. The key words here are "quickly" and "accurately". Your "quick" may be my "dead slow", or visa versa. Ditto, your "reasonably accurate" may be my "intolerably sloppy", and visa versa. For speed, try visualizing how you want it to rise, and then count of seconds. For accuracy, think of the most precise task you want the thing to undertake, and either describe that task to us or figure out the accuracy that it will need to succeed at that task. Sep 13, 2013 at 17:47
• Probably raising from straight down to out 90 degrees in around 1 to .5 seconds maximum and accuracy of at least a centimeter at the end of the arm which means probably at the very least 300 positions for the motor. Sep 16, 2013 at 17:11

So you want to lift 9kg of mass, which means a force of around 90N, that's on an arm that's 64cm long, in one second.

$\frac{\left(90N\right)\left(0.64m\right)}{1s}\simeq58W$

So, that's probably doable.

You don't say anything at all about gearing -- how do you propose to take the high-speed, low-torque power that comes from most motors and turn it into the low-speed, high-torque (or high-force) power that you need for the arm? Do you have space allocated for that, or were you planning on attaching your motor output directly to the arm?

• Preferably coming straight from the motor to the arm or at least having the gearing fit into the space I had allocated for the motor. Sep 24, 2013 at 0:55
• 9Kg is the mass of the robot arm this doesn't include load right? Jan 10, 2014 at 4:17