High torque servo

[mattscott251]

I'm looking for two motors or servos to move about 100 pounds. I used the rmf calculator but i want to double check with someone who actually knows what theyre doing. I found this servo,

http://www.active-robots.com/products/motorsandwheels/futaba-servomotors.shtml

I'm looking at the S3306 Hi-Torque Servo

If I modify this for 360 degree rotation, and I have two of these, one on a wheel of its own, would it work? The robot doesn't need to move fast, under 5 mph is fine.

[hopslink]

Plug the values for torque and speed from the linked page into the motor characteristics section (0.25kg*m and 1.042 rev/s) and you get a motor RMF of 0.260.

Enter the robot mass and number of driven wheels then adjust the other parameters to get the robot RMF down to 0.26 or below. You will find that the robot will only move on a flat surface (<0.6degrees) and will max out at 0.5ish mph if geared for speed. To summarise - very underpowered with those motors.

[definitionofis]

I have a similar weight problem.
I am trying to move 22 lbs. around at very slow speed and potential 27 degree slopes
and I come up with a requirement of two 5 kg-m torque motors (7000 oz-in.).
That is twenty times more torque than those servos.

None of the hobby robot shops have any motor like that, geared down to slow speed.

I am looking at this car air-conditioning baffle mover as a possibility:
It has no torque specifications.  Five RPM gear down sounds promising, though.
Plus the reviews mention somebody needing 15 lb-ft. That is 2880 oz-in.. He got the units wrong.
So maybe he has no clue.
I'm guessing it is in the ball park, but maybe under powered.

http://www.allelectronics.com/make-a-store/item/DCM-276/5-RPM-GEAR-MOTOR-12VDC/-/1.html

If anybody has a suggestion, please contribute.
Reducing weight is not an option.

I just thought of a radical idea.
I built a solar panel rotation device using a satellite linear actuator.
It runs very slowly on low voltage and low current.

I imagine it has a very long threaded rod down the centre.
I could attach a threaded rod to each wheel through a nut and turn the wheels like an old fashioned steam locomotive
by powering a DC motor, that switches polarity at the two extreme positions of the rod.

Direction and torque control without momentum would be difficult. This gets complex.  I could make a career out of it.  
Locomotive engines are more intricate than what meets the eye, now that I think about it.
Ok ...  back to searching for gear boxes and motors.

Getting back to ridiculous ... I could push my robot with a linear actuator acting like a pogo stick. Bwahahaha.
Don't laugh too hard. That is how babies get around crawling on the floor. They don't have wheels, you know.

edit:
I read this link:
http://www.societyofrobots.com/robotforum/index.php?topic=7927.0
and all my torque calculations look too high compared to their discussion.
But I used this page:
http://www.societyofrobots.com/RMF_calculator.shtml


Mass       10 Kg
Desired Velocity       0.01 m/s
Desired Acceleration       0.001 m/s/s
Expected Efficiency    80   %
Incline Angle    27   degrees
Wheel Diameter    0.22 m   
# of Powered Wheels     2   wheels

RMF Results: 0.145

I did find bike motors and planetary gears in this thread,
which might suit my use. Wheel chair motors are obviously too strong.
I wonder if I can match that 5 rpm air-conditioner component to the 4.5:1 planetary gear down ?
http://endless-sphere.com/forums/viewtopic.php?f=28&t=9604

I stared at the planetary gear long enough to figure the 12mm (almost 1/2") shaft size is turning the 0.32" shaft 4.5 times faster.
So the gearing is in the wrong direction (speeding up), as far as my idea of matching it up with that 5 rpm motor.

Drop that idea and start looking for 0.32" motor shafts. Those bore sizes are all actually metric: 8mm and 12mm.

edit: Now I understand why they have two gear ratios for that planetary gear.
If I mount the wheel right through those planetary gear assembly holes, using it like a wheel hub, and ignore the 12 mm bore hole,
and mount a 8mm shafted motor to that smaller bore, I think I get that slower ratio.
I've never used planetary gears. I'm thinking out loud.

[waltr]

For low cost motors try an automotive junk yard. Windshield wiper motors and window opening motors come to mind.  No idea what the speed or torque is.

[definitionofis]

That is a good idea. Thanks. Although, I wish I could do some calculations first.

edit:
I found a couple 12V 3.0A (no load)  100 rpm  12 pitch   9 tooth  0.75" diameter gear (overall size: 3.5"x7"x1.75")
geared power window motors for $30 each from
a tiny auto parts scrap yard only half a mile from my house. I didn't know I had one in my neighbourhood.
They feel like high torque motors.  The wiper motors were bigger and stronger; too big.
I like these power window motors.

I'll see if they can move my 22 lb. robot around by trying them.

I am going to need the current regulator to go with these.
I measure one ohm across the terminals while stationary.
I suppose they are drawing less than 12A at immediate start up
since inductance would be higher than one ohm if I pulse them.

I`m not sure about the characteristics of DC motors.
Putting my ammeter in series might not be telling the whole story.
Now I see 2.7A, not 3.0A, after they warmed up (or maybe my
cheapo multi-meter warmed up.   ).

Something similar is here, if anyone else is looking for higher than
hobby-shop torque:
http://www.car-stuff.com/boschwindowmotor.htm

Here is a guy who discussed the same high current issue, regarding
a high current H-bridge for a power window motor (discussion is circa 2005).
http://www.electro-tech-online.com/robotics-chat/17113-need-help-high-current-h-bridge.html

Well, there`s a 10 amp one:
http://www.criticalvelocity.com/item.php?itemid=hb03

and a 15 amp one:
http://www.robotshop.ca/critical-velocity-pwm-motor-speed-controller-1.html

and a $30, 9 amp one:
http://www.robotshop.ca/pololu-high-current-motor-driver-2.html

One motor control for $28 , 6 amp continuous duty, 25 amp peak surge.
http://www.robotshop.ca/solutions-cubed-simple-hbridge.html

I`ll go with this $54 two motor 9 amp control on one board:
http://www.robotshop.ca/pololu-dual-motor-driver-9a.html

[SMSimon]

Were you able to modify the Futaba S3306 servo for continuous rotation?  My son would like to do something similar.  Do you have any advice?
Thanks, Sandy