1
$\begingroup$

I have a class project where I am using several servos to control a telescope. I'm trying to figure out the torque needed to rotate the tilt of the telescope with a servo from 0 to 90 degrees. I haven't come across any similar problems online that have examples. I know that the one I am currently using is close, but just isn't enough to lift the telescope up by itself.

I currently have an HS-311 (51oz-in) that isn't working. The servo I would potentially purchase in place is the HS-645MG (133oz-in).

From the experiments I've done, I've been able to rotate a ~6oz object almost perfectly, but the ~12oz telescope is too much. I assume that if I get the 133oz-in servo, it will be able to rotate 2.6 times the mass or 15.6oz in this case.

Here is a modified diagram of the original setup. The circular disc on the servo tip is 1" in diameter.

setup

And here is what it looks like assembled (without telescope attached yet).

assembled

$\endgroup$
  • $\begingroup$ Can you add a counterweight so that the telescope is balanced? You can reduce the torque to any arbitrarily small amount (ultimately the friction in the bearing) $\endgroup$ – Martin Beckett Oct 18 '13 at 4:48
  • $\begingroup$ The telescope is pretty balanced already, the problem is when the telescope is completely vertical at 0 degrees (eye piece towards the ground, front towards the sky) and trying to move to 90 degrees as seen in this image. $\endgroup$ – flip66 Oct 18 '13 at 5:00
  • 1
    $\begingroup$ Right, which is very unbalanced. The typical way to solve this problem is a counterweight, as Martin suggested. $\endgroup$ – Colin K Oct 18 '13 at 5:33
  • $\begingroup$ In which case the telescope would just tip forward after it hit the horizontal mark (90 degrees). $\endgroup$ – flip66 Oct 18 '13 at 5:37
  • $\begingroup$ No, it would not. I think you might be confused about something. A counterweight would result in the load remaining stationary in any position unless an external force were applied. That way large servos are not required. All the servo needs to overcome is bearing friction and the inertia of the telescope. $\endgroup$ – Colin K Oct 18 '13 at 15:57
0
$\begingroup$

counterweights only help the servo in one direction if the telescope is already mounted on its cg. also, telescope completely vertical is when torque required to turn it is maximum.

you can either bring the scope closer to the axis of rotation by shortening the armature (it already looks as short as it can be), add gears to decrease the gear ratio (telescope slows down by this ratio, but torque increases by this ratio).

enter image description here

starting from scope being completely vertical, activate the servo so it wants to go anticlockwise. it shouldnt have the torque to accomplish this. now manually raise the scope until the servo does has torque to move the scope on its own. the angle where this happens is the critical angle. Do $51\over\cos\theta$ to calculate what torque you need for a new servo that can do this on its own.

my advice would be to use gears. specifically worm gear for the driver (servo). which will protect the system from failing like above (if the critical angle is reached it wont fall over, just stall in position). worm gears have the advantage that they can only be driven from the worm, and for example, cannot be forced to turn by pushing on the telescope, bumping into it etc. the disadvantage is the very low gear ratio (slow but high torque). and you would need a motor not servo..

$\endgroup$
  • $\begingroup$ ok so say my angle is about 20 degrees, i would get 51/cos(20) = 124.97. Would that mean that a servo with torque 125 or greater would have a stall angle of 0? $\endgroup$ – flip66 Oct 18 '13 at 17:20
  • $\begingroup$ no, you used radians in your calculation... you actually only need 54 oz/in. you probably want to add 20% to that as a safety margin. if i back calculate, it shows me that the cg of the telescope is mounted 4.5" off the servo axle. is this accurate? $\endgroup$ – gregsan Oct 18 '13 at 18:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.