### Author Topic: Robotic arm  (Read 807 times)

0 Members and 1 Guest are viewing this topic.

#### alehandro

• Beginner
• Posts: 5
• Helpful? 0
##### Robotic arm
« on: March 01, 2014, 03:01:01 PM »
Hi new here, I'm making a project that need's a robotic arm. My base will rotate 180 and the elbow will have a length of 15cm and weight 1kg. Which should the torque of the elbow servo be? Is this a good one?? http://www.ebay.co.uk/itm/1-2-4x-Digital-MG996R-High-Torque-Servo-Metal-Gear-for-RC-Car-Truck-Boat-Model-/400669823345?pt=UK_ToysGames_RadioControlled_JN&var=&hash=item5d49c85171 Thanks..

#### bdeuell

• Full Member
• Posts: 84
• Helpful? 3
##### Re: Robotic arm
« Reply #1 on: March 01, 2014, 05:03:20 PM »
torque = force*distance (distance being the radial distance between the axis of rotation and the load vector)

I'm not sure I understand your design ... you talk about the base and then the elbow. Is there more than one joint? Perhaps a better description or ideally a picture.

If you are designing a joint that will move a 15cm long arm in a vertical plane then the max load would be when the arm is in the horizontal position (where the radial distance the force vector acts on is longest). This would give a torque of 15 kg-cm not accounting for any acceleration, frictional losses, or the mass of the arm. As your servo can only achieve this torque at stall it is probably a little too small especially given the other loads not accounted for in the calculation.

Note N-cm are the correct units to express torque but I used kg-cm because that is how your servo is speced. To convert mass to weight (force) use Force = mass*acceleration (acceleration in this case is acceleration due to gravity = 9.8 m/s^2)

#### alehandro

• Beginner
• Posts: 5
• Helpful? 0
##### Re: Robotic arm
« Reply #2 on: March 03, 2014, 06:30:20 AM »
Thank you very much for your answer!! To be more precise my robotic arm will have a base that will only rotate 180 degrees and then 10cm arm L1 until the elbow ant then another 15cm long arm L2. That's it only that. My L2 will be 1Kg with the arm mass. The L2 will be at a vertical position and I want it to move 45degrees right and left. I mean that the center will be at the vertical position and wont go at a horizontal position ever.

Also what does at stall mean? It's probably a stupid question but my inglish is a little rusty..

And to find the exact torque for my servo can you explain it one more time . You wrote me that "torque = force*distance" so my distance is 15cm and my force is "mass*acceleration" which mass in this case is 1kg and acceleration 9.8m/s^2?? Thank you..

#### jwatte

• Supreme Robot
• Posts: 1,343
• Helpful? 81
##### Re: Robotic arm
« Reply #3 on: March 03, 2014, 11:28:46 AM »
You will need to post a drawing or sketch of some kind -- I don't understand what you intend to convey with "L2 at vertical."

If the longest distance your gripper will be from the center is 25 centimeters, and the weight of your load plus the arm is 1 kg, then the holding torque of the L1 servo needs to be at least 25 kg.cm.

"stall torque" is the rating for an actuator/motor at which the motor stops and cannot produce any more torque. A motor that is held in stall for more than one or a few seconds is likely to destroy itself. When you see only one rating for a motor, it's usually the "stall torque," if you intend to apply continuous torque, the "working torque" of a motor will typically be 0.20 or 0.25 times the stall torque.

Thus, for a holding torque of 25 kg.cm, you may need a motor with stall torque rating of 125 kg.cm.

#### alehandro

• Beginner
• Posts: 5
• Helpful? 0
##### Re: Robotic arm
« Reply #4 on: March 03, 2014, 12:39:53 PM »
Thank you for replying. Here's a sketch, if it's to sloppy tell me to do a better one.. I will have a sun panel instead of a gripper and some other stuff too and that's why the weight of the arm and load will be 1kg.. It will go only 45degrees left and 45degrees right and the center of the L2 arm 15cm will be facing up. Which should the torque of the elbow servo be?
« Last Edit: March 03, 2014, 03:46:32 PM by alehandro »

#### bdeuell

• Full Member
• Posts: 84
• Helpful? 3
##### Re: Robotic arm
« Reply #5 on: March 03, 2014, 05:31:29 PM »
for servo 2 the distance used to calculate torque would be cos(45deg)*15cm=10.6 kg*cm because you are limiting the range of motion. I suspect acceleration is of little concern (if I am interpreting your use correctly) you will be fallowing the sun with the arm. As Jwatte mentioned you do not want to operate the servo continuously at stall so you will need to use a larger servo to prevent damage due to overheating.

Servo 1 can be smaller than servo 2 but it is difficult to specify an exact torque without knowing the details of the design as much of the load may come from friction on this axis.

#### jwatte

• Supreme Robot
• Posts: 1,343
• Helpful? 81
##### Re: Robotic arm
« Reply #6 on: March 03, 2014, 06:51:07 PM »
Ah, this is for a solar facing panel!

Why do you need the 10cm arm? Can't you just mount the panel on a bracket to the servo horn, say 2 cm out? If so, you'd need only 2 kg.cm holding torque, so about 10 kg.cm stall torque.

#### alehandro

• Beginner
• Posts: 5
• Helpful? 0
##### Re: Robotic arm
« Reply #7 on: March 04, 2014, 06:59:59 AM »
Yea it's for a solar facing panel..
It would be simpler like that but I'm doing a project that a professor gave me and it must be exactly like that!! Like the sketch I mean.. So like bdeuell said "for servo 2 the distance used to calculate torque would be cos(45deg)*15cm=10.6 kg*cm" is this my holding torque?? and then is this servo good for my case http://www.ebay.co.uk/itm/1-2-4x-Digital-MG996R-High-Torque-Servo-Metal-Gear-for-RC-Car-Truck-Boat-Model-/400669823345?pt=UK_ToysGames_RadioControlled_JN&var=&hash=item5d49c85171 because the stall torque is 15kg.cm. So stall torque is 15kg.cm so  the holding stall torque is: 15kg.cm*0.2=3kg.cm?? From what jwatte wrote me or am I understading it wrong?
« Last Edit: March 04, 2014, 07:52:48 AM by alehandro »

#### bdeuell

• Full Member
• Posts: 84
• Helpful? 3
##### Re: Robotic arm
« Reply #8 on: March 04, 2014, 10:14:59 AM »
That servo would probably move your load but would quickly burn up. Your calculations are correct. From the estimation of the holding torque Jwatte provided you have a 3kg-cm  holding torque but you will need to hold at least 10.6kg-cm. While the servo is capable of providing this torque (because it is less than the stall torque) constantly supplying this torque will cause the servo to overheat.

#### jwatte

• Supreme Robot
• Posts: 1,343
• Helpful? 81
##### Re: Robotic arm
« Reply #9 on: March 04, 2014, 11:14:51 AM »
Also, the amount of holding torque versus stall torque varies by motor and implementation -- better cooling increases the 0.2 factor.

I'd start looking at a 40 kg.cm servo, and measure it under actual maximum load (arm angled 45 degrees.) Use a contactless thermometer and measure how warm it gets. If it's less than 60 degrees C on the case after 1, 5, and 15 minutes, you're probably OK.

Also, if you're doing this for a project -- for how long does the load need to be at 45 degrees? You can get away with a weaker servo, if the time at max load is limited (a dozen seconds, say.) I'd look at a 20 kg.cm or higher servo, still.

These servos might do it:
http://www.robotshop.com/en/herkulex-drs-0201-robot-servo.html
http://www.trossenrobotics.com/dynamixel-mx-28-robot-actuator.aspx

These servos almost certainly would do it:
http://www.robotshop.com/en/herkulex-drs-0402-smart-robot-servo.html
http://www.trossenrobotics.com/p/mx-64t-dynamixel-robot-actuator.aspx (I use these for my walker robot)

Cheap crap that might get you through the assignment if the demo is short:
http://www.hobbypartz.com/33p-solarservo-d772.html (I use some of these for four-wheel steering on my rover)

An alternative would be to use a stepper motor (or brushed DC motor) with appropriate controller, and a limit switch to sense the "zero" position. That would probably be cheaper, but more fiddly to get to work.

#### bdeuell

• Full Member
• Posts: 84
• Helpful? 3
##### Re: Robotic arm
« Reply #10 on: March 04, 2014, 05:04:08 PM »
If you do move away from using a hobby servo you may want to consider using a worm gear to drive. This would allow you to support your load without keeping the motor powered.

#### alehandro

• Beginner
• Posts: 5
• Helpful? 0
##### Re: Robotic arm
« Reply #11 on: March 05, 2014, 02:14:39 PM »
Oh that sounds nice.. Will the worm gear be connected to the servo so that the torque gets higher??

#### bdeuell

• Full Member
• Posts: 84
• Helpful? 3
##### Re: Robotic arm
« Reply #12 on: March 05, 2014, 03:39:57 PM »
Yes the worm gear would work to increase your output torque.

Backlash can be a concern in some applications but I suspect your arm will not be going completely vertical and your loads will be static. This means the torque on the drive will always be in one direction and the backlash will probably not cause a problem.
« Last Edit: March 05, 2014, 03:43:51 PM by bdeuell »

Get Your Ad Here