Author Topic: DC Motors for hand: small AND high torque?  (Read 4330 times)

0 Members and 1 Guest are viewing this topic.

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 375
  • Helpful? 10
DC Motors for hand: small AND high torque?
« on: January 04, 2011, 10:36:33 AM »
I'm trying to build a robot hand (although at the moment, I'm just working on a single finger).  Each finger will have three joints, although in theory I could reduce that to two joints to make my life easier, but we shall see.

My biggest problem I'm coming up against is the size of the motors I'm using.  I can't seem to find a small enough motor, with a gearbox, to actuate the hand.  It doesn't have to be the size of a human hand.  I have a plan the each finger can be 1" wide and probably 6" long, but it's remarkably hard to find a motor that will fit into the links and drive the joints.  Any thoughts on a good one?

Theoretically I'm also open to other methods of actuation, but the restriction is that the hand has to be fully actuated -- I don't want a single motor driving all three links of the fingers, as is popular in the hobby robotics websites.  My thoughts go that the simplest method of actuation is one motor per joint, with the downside being the motors may never be small enough to work. 

MIKE
Current project: tactile sensing systems for multifingered robot hands

Offline waltr

  • Supreme Robot
  • *****
  • Posts: 1,944
  • Helpful? 98
Re: DC Motors for hand: small AND high torque?
« Reply #1 on: January 04, 2011, 12:09:32 PM »
Expensive but these will fit your requirements:
http://www.micromo.com/?gclid=CMu57-HtoKYCFVln5QodsB1Rnw

Check their "Application Case Studies" and "Technical Library".

http://www.micromo.com/n280126/n.html

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 375
  • Helpful? 10
Re: DC Motors for hand: small AND high torque?
« Reply #2 on: January 04, 2011, 01:45:02 PM »
Hm, this seems perfect.  A bit of a concern that pricing information is relegated to "request a quote".  I wonder if anyone has any information on an estimated cost for one of their motors/encoders/gearbox combos? 

I was trying to keep the hand under a budget of $1000, but it looks like I'm going to have trouble doing that. Oh well, that's what grant requests are for :-P

Thanks for the reply,

MIKE
Current project: tactile sensing systems for multifingered robot hands

Offline waltr

  • Supreme Robot
  • *****
  • Posts: 1,944
  • Helpful? 98
Re: DC Motors for hand: small AND high torque?
« Reply #3 on: January 04, 2011, 02:06:23 PM »
Some times some of these motors can be found surplus. Look for surplus sellers and even eBay if you are flexible in the motor/gear box selection.

Here is another maker of small motors and they do have prices listed.
http://www.maxonmotorusa.com/production.html

And another:
http://www.portescap.com/BrushDC
« Last Edit: January 04, 2011, 03:16:02 PM by waltr »

Offline azy

  • Full Member
  • ***
  • Posts: 107
  • Helpful? 0
Re: DC Motors for hand: small AND high torque?
« Reply #4 on: January 05, 2011, 01:39:01 AM »
Most fully articulate hands you see have fixed wrists. Control rods for each finger joint run into the arm where there is plenty of room for actuators. Fixed wrist hands are simple and cheap because all the motors are located in the arm.

A nodding donkey type mech means that the motors only have to be switched on and off and reversed. Rotary motion is turned into an up and down joint movement. You will prolly want to gear the motors down as the joints will be moving very fast. But you will be able to use very cheap and small motors (less than 5 bucks each) 15 joints x 5 digits = 75



It is possible to have an articulated wrist , but the control /rods wires have to run exactly through the wrist joint in order to avoid moving the wrist when you don't want to.

You have the additional option of steadying the wrist with powerful actuators that lock it or counter the unwanted forces generated by the finger control rods

how much articulation do you need & what kind of strength do the fingers have to have ?

Im in the process of building a fully articulated hand/wrist & arm myself
« Last Edit: January 05, 2011, 01:54:13 AM by azy »

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 375
  • Helpful? 10
Re: DC Motors for hand: small AND high torque?
« Reply #5 on: January 05, 2011, 08:24:34 AM »
I guess it's a matter of my learning how to make the rod/wire control mechanism.  Issue is we're using a robot arm already -- a Fanuc F5, so we don't have much room for actuators in the body of the arm.

I'm thinking of having a 3 fingered hand, similar in spirit to the Barrett hand.  In fact, I'm probably just going to make my own version of the Barrett Hand (http://www.barrett.com/robot/products-hand.htm).  I'm debating whether each finger needs to be 3 or 2 link.  The former would provide redundancy, the latter would simplify the design.  In all honesty, this hand isn't supposed to do anything nearly as complex as a human hand, so I'm thinking of a simpler prototype rather than a more complex one.

I have to go through the math to figure out the strength I need from the fingers.  My guess is that if I can get 1/2 a newton out of each one as a max, as measured when that finger alone is pushing on a fixed force plate or something, I'd be pretty happy, but that is just an estimate and it might be way too low or way too high.  I'll try to do the calculations and let you know today.

Any advice for a novice hand builder? :-)

MIKE
Current project: tactile sensing systems for multifingered robot hands

Offline mstachoTopic starter

  • Supreme Robot
  • *****
  • Posts: 375
  • Helpful? 10
Re: DC Motors for hand: small AND high torque?
« Reply #6 on: January 05, 2011, 10:12:33 AM »
As an update, I did some preliminary joint torque calculations as follows:

Assume the fingers have 3 links, with dimensions 70mm, 50mm, and 30mm (this is approximately the size I want, although the 30mm link might be a bit longer to accommodate its motor/gearbox).  I'm also restricting the joints to move in the range of

[pi/4, 0, 0] to [3*pi/4, pi/2, pi/2] radians.

I tried two maximum forces at the finger tip: 5N and 3N, and ran it through an optimization program to find the joint angles that could provide that force for which the torque would be a maximum. 

Turns out that 5N would require a maximum joint torque of 0.75Nm (that's 106.2087 oz-in to anyone using that measure), and the 3N would require 0.45Nm, or 64oz-in, of torque as a maximum. 

The other motors also supplied a torque, but they were all smaller than this maximum, and I want to design to the maximum constraint.  Luckily, cheap motors from robotshop can provide this torque.  They're a bit big, but that's not a big deal for now. 

Once the prototype is done, I'll probably invest in higher torque motors, but for now it doesn't have to do much...

MIKE
Current project: tactile sensing systems for multifingered robot hands

 


Get Your Ad Here