Servo power /torque /speed
11-12-2004, 02:37 PM
#1
Member
Thread Starter
Join Date: Nov 2003
Location: Coyonkulma, FINLAND
Posts: 74
Likes: 0
Received 0 Likes
on
0 Posts
Servo power /torque /speed
Hello
I have tried to carefully compare different servos. Main parameters are trivially TORQUE and SPEED. I noticed there is a relationship between these. Thats because I can vary the servo ARM LENGTH. Pick up a servo and then make your setting for desired speed or torque.
I made a calc sheet http://kkweb.mine.nu/~jyrki/servo/ for this.
Then I noticed if I reach for speed and use a narrow angle with long servo arm, I dont get a good accuracy because of the electrical resolution of the radio system is not utilized. In the other hand, I get more mechanic tolerance if the arm is short. So the thumb rule seems to be a setup with a large travel adjust and the best servo selection for the use.
However there seems to be great benefit for the speed if the arm is long. If nothing else, my sheet proves that you can ruin your top of the art tail servo speed with a too short servo arm!
Because torque/speed- relation CAN be converted to favour one or the other, I wondered what is the general quantity for the "goodness" of a servo. I wonder would that be the net POWER (P) of a servo. That would be TORQUE (Nm)per SPEED (m/s)?, measured as Watts? Please check before you trust! As a curiosity when you know the voltage of the battery supply, you can calc current (I) =P/U. That would be a greatly fictional current thought
).. and therefore must be handeld as a theoretical 100% efficiency under a 100% duty. A curiosity in my sheet.
Jyrki
ps. How did I calc the power? Well.. I noticed 1cm arm makes a 1cm distance at 60 degrees. How convinient! the corresponding arc is a bit longer, but close enough. Well.. correct or not...I converted the arm to 1m distance. 10kg/cm would make 1 Newton meter torque. If it travels it in 1s it makes 1 Joule work with 1 Watt.
This comes from my "law of the squirrel"
A squirrel weights 100g. It causes 1N force. (ok... 0.981N but the squirrel maybe lied the weight). Well.. it climbs a meter. It makes 1 Joule work. It takes 1 second and Ws=J, so it has a 1W engine. It has a simple 1V power supply,P=U*I, so it consumes 1A. U=R*I which tells the internal resistance is 1 ohm.[>:]. I simply applied this to a 1 meter fictional servo arm.
This is very practical rule! For example give a peanut to a squirrel and look how high it climbs. Each peanut Joule makes a meter. This assumes the squirrel dont loose anything, but practically they do.[:-] In the scool maths we just dont want to think about it. I just wonder how this works when you dont calculate with SI-units!!?? must be a nightmare. A squirrel with square feets ?? I dont want to think more about such an animal.
I have tried to carefully compare different servos. Main parameters are trivially TORQUE and SPEED. I noticed there is a relationship between these. Thats because I can vary the servo ARM LENGTH. Pick up a servo and then make your setting for desired speed or torque.
I made a calc sheet http://kkweb.mine.nu/~jyrki/servo/ for this.
Then I noticed if I reach for speed and use a narrow angle with long servo arm, I dont get a good accuracy because of the electrical resolution of the radio system is not utilized. In the other hand, I get more mechanic tolerance if the arm is short. So the thumb rule seems to be a setup with a large travel adjust and the best servo selection for the use.
However there seems to be great benefit for the speed if the arm is long. If nothing else, my sheet proves that you can ruin your top of the art tail servo speed with a too short servo arm!
Because torque/speed- relation CAN be converted to favour one or the other, I wondered what is the general quantity for the "goodness" of a servo. I wonder would that be the net POWER (P) of a servo. That would be TORQUE (Nm)per SPEED (m/s)?, measured as Watts? Please check before you trust! As a curiosity when you know the voltage of the battery supply, you can calc current (I) =P/U. That would be a greatly fictional current thought
).. and therefore must be handeld as a theoretical 100% efficiency under a 100% duty. A curiosity in my sheet.Jyrki
ps. How did I calc the power? Well.. I noticed 1cm arm makes a 1cm distance at 60 degrees. How convinient! the corresponding arc is a bit longer, but close enough. Well.. correct or not...I converted the arm to 1m distance. 10kg/cm would make 1 Newton meter torque. If it travels it in 1s it makes 1 Joule work with 1 Watt.
This comes from my "law of the squirrel"
A squirrel weights 100g. It causes 1N force. (ok... 0.981N but the squirrel maybe lied the weight). Well.. it climbs a meter. It makes 1 Joule work. It takes 1 second and Ws=J, so it has a 1W engine. It has a simple 1V power supply,P=U*I, so it consumes 1A. U=R*I which tells the internal resistance is 1 ohm.[>:]. I simply applied this to a 1 meter fictional servo arm.
This is very practical rule! For example give a peanut to a squirrel and look how high it climbs. Each peanut Joule makes a meter. This assumes the squirrel dont loose anything, but practically they do.[:-] In the scool maths we just dont want to think about it. I just wonder how this works when you dont calculate with SI-units!!?? must be a nightmare. A squirrel with square feets ?? I dont want to think more about such an animal.
11-12-2004, 04:01 PM
#2
Senior Member
Join Date: Sep 2002
Location: Rochester,
NY
Posts: 1,373
Likes: 0
Received 0 Likes
on
0 Posts
RE: Servo power /torque /speed
All the math is fine and dandy for a quick comparison, but you're forgetting that the speed of a servo is no-load, and the torque is stall torque. In the real world, speed and torque follow a varying curve dependant on load which you can only calculate by basically dyno'ing the servo's and I don't know anyone that does that with a large number of servo's for comparison. Considering the difference in servo motors, the way they're geared, and any mechanical advantage that may be gained or lost from linkages, include travel distance, speed and resolution and all that fun stuff and it is for all 'sane' purpoes impossible to directly compare two similar servo's relative usefullness for a particular application.
