Servo torque

[drinu]

Hi,

I am planning to do my first robotic arm. With reference to the link below, I want L4 and L5 to be approx 30cm, and L3 approx 6cm. I want the arm to be very precise and the gripper to lift at least 1kg (preferably more).

http://www.robotshop.com/content/images/learningcenter/robot-arm.jpg

Question1)

What are the approximate torque values (in kg-cm) required by every motor to lift that kind of load and keeping the precision? (just to have a rough idea)

Question 2)

I was watching the video (link below) and at 8:00 minutes he said that when for example you have a 20kg-cm servo, that means that if you extend the arm by 2cm from the shaft, the torque will drop by half. Does that mean that if the arm is extended by,  lets say 7cm, the servo torque will drop from 20 to less then 1 kg-cm ??

Microcontrollers - Introduction to Servos and Understanding Torque

Question 3)

I was also considering using high torque DC motors (like the one attached in the link below, where I will attach a pot at every joint to sent analogue data to monitor the arm position. Will this work? will they be sensitive to different loads at the gripper?

http://www.ebay.co.uk/itm/DC-24V-0-33A-150RPM-10-8kg-cm-High-Torque-Gear-Box-Motor-/380532858650?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item5899869f1a

http://www.ebay.co.uk/itm/12V-DC-HIGH-TORQUE-30-RPM-Reversable-Motor-GBox-ARDUINO-RASPBERRY-Pi-/140909963312?pt=UK_BOI_Electrical_Components_Supplies_ET&hash=item20cee36830


Thanks in advance.

[jwatte]

The load supported drops by distance. The torque is the same.

[drinu]

Thanks for the reply.

So is it correct that if having a 20kg-cm servo, the load that the motor can handle will drop below 1kg if the arm is 7cm long?? so to do the robotic arm with the dimensions mentions previously, I will be needing servos with very high torque.

[jwatte]

What do you think "kg" and "cm" mean in torque? :-)

A load of 1 kg at a distance of 7 cm from center of rotation needs 7 kgcm to stay in place when posed 90 degrees horizontally, and more than that (I'd say at least 10 kgcm) to get movement more than just counteracting gravity.

[drinu]

Is there a way to change the motors position (instead of at every joint) so to increase the torque?

[waltr]

Is there a way to change the motors position (instead of at every joint) so to increase the torque?

No, Torque is measured from the center of rotation.
To increase the torque output of a given motor use speed reduction gearing.

Look up Torque in Wiki and other Web sources for definitions and examples.

[jkerns]

Is there a way to change the motors position (instead of at every joint) so to increase the torque?

Use a lever.

[drinu]

Hi, thanks for the replies.

Will the attached sketches help in any way increase the load capacity?? (when compared to the motors being attached to each joint/axis)

[jwatte]

You will increase torque if you build a system with mechanical leverage -- one that turns fewer degrees at the load than it does on the motor. For example, if your load-holding piece turns 45 degrees when your motor turns 90 degrees, you will get a 2:1 increase in torque around the pivot point of the load-holding piece.

[jkerns]

Hi, thanks for the replies.

Will the attached sketches help in any way increase the load capacity?? (when compared to the motors being attached to each joint/axis)

So you are using motors to turn a screw to make a linear actuator? Likely you would get more force than a direct drive motor - but of couse, you are trading off speed. You really need to do the math to translate the motor torque into the linear force via the pitch of the threads to translate from rotary to linear motion. I would assume that the efficiency of your screw drive is only about 50% - there tend to be a lot of friction losses.

[drinu]

Hi, thanks for the replies.

Another question:

If I control the motor speed using PWM, when reducing the speed will the torque also decrease?

[jwatte]

Torque in an electric motor is a function of how much current you can drive into it. The good thing with PWM is that it can use a high voltage for the times it is on, and thus have higher torque available even at slow speeds than a voltage-based alternative. However, at some level, you have to reduce the torque output to run slower, because a higher torque will cause acceleration up to equilibrium.