Servo motors[]
General info: http://www.horrorseek.com/home/halloween/wolfstone/Motors/svoint_RCServos.html
Not sure where this info hails from (but it needs cleaning up hint hint nudge nudge):
-typically 4.8 - 6 VDC to power with higher power giving faster movement and more torque
- almost all modern servos have positive power in going to the middle connector
- more generally the dark coloured wire is ground, the middle one is power and the other colours are the signal
-servos are controlled by a series of pulses where the length of the pulse indicates the position the servo should take (i.e. the angle)
-the range of the servo is generally from 0 - 180 degrees with a corresponding pulse width range of (0.8 ms to 2.5 ms)
-increasing the pulse width usually by about 10 microseconds results in one degree of movement
-the rate at which the pulse is sent isn't too important but typical rates are 400 Hz (with 2.5 ms pulse spacing) and 50 Hz (20 ms pulse spacing)
-the driving pulse is usually specified as 3 -5 V peak to peak
FOR GENERATING A PULSE WITH FUNCTION GENERATOR
-use square wave
-power supply approximatlely 4.8 V (or to match power supply of servo)
-adjust width of pulse by adjusting the FREQUENCY SIGNAL
-any pulse width narrower than the neutral point (the point at which the servo stays at 90 degrees) will cause the servo to move to a position less than 90 degrees and vice versa
-generally the neutral point is around a pulse width of 1.52 ms
General Information[]
- How a servo motor knows it's position, PWM,
Here's some info about sail servos, torque, rotations, sheet travel, etc: http://www.joliebrisemodels.co.uk/tenth/servo.html
Connectors[]
Servo connectors are mysterious
http://www.hansenhobbies.com/products/connectors/servoconnectors/ supplier
http://parts.digikey.com/1/parts/270992-conn-term-female-24-30awg-tin-16-02-0096.html This MIGHT be a servo connector (linked from forum)
http://www.servocity.com/html/servo_connectors___pins.html Another supplier
http://www.rctech.net/forum/radio-electronics/251456-servo-connector-name.html The forum post with the most info "It's called a JST connector (or bec connector)."
Modified Servo Motor[]
Position Feedback[]
The servo motor can be modified to feedback data detailing the motors current position. The motor that is modified now has a fourth shorter wire(coloured yellow) exiting with the other three wires(red, black, and yellow). Measuring Voltage drop between the new yellow wire and ground gives a value to be used to calculate the motors position. Further testing is required to determine a formula to convert from voltage to an angle in degrees.
The process for doing this is modification is detailed<a href="http://forums.trossenrobotics.com/tutorials/how-to-diy-128/get-position-feedback-from-a-standard-hobby-servo-3279/"> here </a>. Note that in the tutorial when it describes measuring voltage from the yellow wire to ground this is not referring to the yellow wire that is immediately visible when opening the motor. You must lift off the micro chip and use the yellow wire bellow.
Continuous Operation[]
See instructions: http://www.kronosrobotics.com/an116/GAN116.shtml
Rudder[]
http://www.elemodel.com/shopen/html/product_1196.html rudder servo 4.8->7V
The 2007-2010 rudder servo used is a large Tower Pro 9805BB Double Ball Bearing servo, which Cory was told was probably overkill by someone at Leading Edge Hobbies. They aren't waterproof; in 2010, when there was water in our hull that we had to drain, we doused this servo and busted it. Testing should be done with a smaller rudder servo in high winds, which will save us some weight if it works.
Main sail[]
In 2009 and 2010, main and job sail were both run off of the smartwinch380EH. Prior to 2009 the sails were controlled with h-bridges and DC motors with potentiometer position feedback.
Power use: Max power: 14.2 watts (based on 6V in)
Idle current: 23 mA
Running current: 650 mA
Stall Current (max) 18 A
SmartWinch 380EH, links to features and specifications.
If smartwinch is acting like it's very noisy, making lots of beeps, or has a very small range, change the 7.2V unregulated battery. Presently, Pololu control is very smooth, even with rudder connected, and RC is very noisy.
With a low battery (presumedly) the SW "trills" (rising beeps) when the rudder is connected -> perhaps there's less power available with rudder?
Jib sail[]
Smartwinch 2009, 2010; both sails on one motor. Prior to 2009, a DC motor, Hbridge, and potentiometer was used to set the jib position separate from the main.
Jib servo DB9 uses Pins 1, 2 and 3. Pin 1 is signal, pin 2 is ground and pin 3 is 5V.
Running with arduino[]
http://arduino.cc/en/Reference/Servo or use Pololu
Possible Future Servo Motors[]
Considerations when sizing motors:
- holding torque
- speed
- distance (degrees usually; some are also continuous)
- type of gears (Motors with plastic gears inside are more likely to break at high force.)
- available servo horns (drum-mountable?)
list of servo motor specs available through RobotShop
http://www.robotshop.ca/robotis-ax-12-smart-servo-1.html Digital servo, has position and temperature feedback
http://www.robotshop.ca/hitec-hs-765hb-servo-motor.html Has a sail control horn (big horn)
http://www.robotshop.ca/servocity-spg785a-45-servo-gearbox-3-1-kit-3.html Servo gearbox, fits specific servos only
Check out this site! Here's some info about sail servos, torque, rotations, sheet travel, etc: http://www.joliebrisemodels.co.uk/tenth/servo.html
Sample RC Boats[]
trimaran with 1260sq in of sail uses the hitec 785 for both main and jib "Inside the main hull, the Trimaran is outfitted with a powerful Hitec drum winch servo. This RC servo will have the power to haul in your sails in any wind."
1200 sq in of sail on a monohull, again, using the hitec 785 for both main and jib races in the marblehead class, one of the biggest rc boats.
our jib is almost exactly .90m^2, roughly the same size. considering that there is much less friction in a jib system than a main and jib, or especially a full rotating mast (as in the mono above), we're still in the clear.
semi related discussion: http://www.theamya.org/hints/sailservo.php
and also http://www.rcuniverse.com/forum/m_3839448/printable.htm
Sail Motor System Design[]
The way to make the control system the simplest is to have the same travel on the servo shafts, but put different size drums so that the resulting distances are different (ie get two 180 degree servos, a large drum for the main and a smaller for the jib); this way they can both get the same commands from the servo controllers but the sails will move the appropriate distance. This can also be accomplished with pulleys to change the ratio by an integer amount.. but I'm not sure if it could do a 5/6 ratio
Motor Sizing without gears or pulleys[]
Adding pulleys makes it more complicated. Here's the calculations for no pulleys.
For a system with a 1:1 ratio (no pulleys):
T = F*r (1) d_max = r*theta_max (2) <- is this right? theta in radians? yes because circumference = 2 pi r = d for a full circle Eliminate r T = F * d_max / theta_max T * theta_max = F * d_max Where: T = motor torque spec (N.m) theta_max = rotation of motor max, in radians (typically Pi) d_max = desired distance (m) F = desired force (N) r = desired drum radius (m) From this solve for torque, then solve for the radius
For speed calculations, 1:1 system
v = d / t (1) v = w * r (2) Solve for w w = v / r = d / t / r Where: w = angular rotation speed motor spec (rad/s) t = 1 s desired system speed (s) d = desired system distance, from desired specs (m) r = drum radius, from the torque calculations (m) v = velocity of sheet (m/s)
The desired specs for 2011 are:
Jib travel: 1.02m ; jib sheet tension: 0-35N ; speeds: full travel in <1s; drum Main travel: 1.15m; main sheet tension: 0-60N; speeds: full travel in <1s; drum
Plug these in, and spec a motor; for theta_max = Pi:
Jib torque = 11.36 N.m radius = 0.3245 m speed = 3.14 rad/s Main torque = 21.96 N.m radius = 0.366 m speed = 3.14 rad/s
These would be huge drums, with huge torque. We will likely need to either find the slipped decimal in those calculations, or gear things up/down. I forget what the basic gear and pulley system equations are; pulley systems are always 1:1 unless there's a moving pulley, or some weird angles are thrown in.
Both the torque and the radius are inversely proportional (directly) to Theta_max; so for a larger theta_max, they drop proportionately. Speed is proportional to theta_max; a motor with more turns must turn faster, proportionately.
About the maximum for a normal, un-geared servo torque = 20 kg.cm = 20*(9.8N/kg)(1m/100cm) = 1.96 N.m.
http://www.robotshop.ca/hitec-hs785hb-servo-motor.html This motor turns 7Pi, slowly (1.32 s for 60 degrees), at 13.2kg-cm = 1.29N.m .
For theta_max = 7Pi:
Jib torque = 1.6 N.m radius = 0.046 m speed = 22.2 rad/s Main torque = 3.14 N.m radius = 0.052 m speed = 21.98 rad/s
Gears[]
Where two gears meet
F1 = F2 -> T1 r2 = T2 r1 d1 = d2 v1 = v2 -> w2 r2 = w1 r1 T2/T1 = w1/w2
Internal to the gears themselves:
F = T / r d = r * theta v = r * w Therefore F = T * w / v
Law 1: Gears increase torque at the expense of speed, and vise-versa.
T * w = constant for a gear system (Fr * v/r = constant)
For two gears on one axle,
w1 = w2 -> v1 r2 = v2 r1 T1 = T2 -> F2 r2 = F1 r1 theta1 = theta2 -> d1 r2 = d2 r1 v2/v1 = F1/F2 = d2/d1 F*v = constant F*d = constant
Law 2: Distance can be increased at the expense of torque
T * d = constant
Law 3: Speed and distance are proportional
w / d = constant
Sizing a Motor[]
For our 2011 specs, then, to size a motor we just need to find one with the appropriate T*w = F*v, and then gear to get them in the correct ratio of T/w. Also, we need to find the appropriate F*d = T*theta
Specs again
Jib travel: 1.02m ; jib sheet tension: 0-35N ; speeds: full travel in <1s; drum Main travel: 1.15m; main sheet tension: 0-60N; speeds: full travel in <1s; drum
Calculation of desired F*v, F*d
v = d_max / t F*v = F * d_max / t where: t = 1s, desired speed of motion d_max = desired motion distance F = desired force Jib: F*v = 35.7 N m/s F*d_max= 35.7 N m Main: F*v = 69 N m/s F*d_max = 69 N m
Some motor stats - also look into http://www.gpdealera.com/cgi-bin/wgainf100p.pgm?I=FUTM0099
Smartwinch 380 EH T = 30kg.cm = 2.94 N.m w = 3.5 rev/s = 7pi rad/s theta_max = 18Pi rad (9 revs) T*w = 64.7 N.m/s T*theta = 166 N m rad HS-785HB (sail winch with drum) T = 13.2kg-cm = 1.29N.m w = (1.38s/60 degrees) = 0.12 rev/s = 0.75 rad/s <- not sure I did this conversion right theta_max = 3.5Pi rad (630 degree rotation) T*w = 0.97 N.m/s T*theta = 14.18 N m rad HS-765HB (sail winch with arm) T = 11kg.cm = 1.078 N.m w = (0.23s/60 degrees) = 0.724 rev/s = 4.55 rad/s theta_max = 0.388 Pi rad (70 degrees) T*w = 4.9 Nm/s T*theta = 0.4 N m rad
To convert from the data sheets to a Nm/s number quickly:
T*w = T_s*9.8/100 * (2*pi/6*w_s) = 0.103 (T_s/w_s) where T_s = torque from spec sheet (kg.cm) w_s = speed from spec sheet (s/60 deg)
-> there's no motors on RobotShop that can meet this
4m Boat Rudder[]
For the 4m boat rudder, a minimum torque of 45Nm was required (factor of safety included of course). As such high torque and low speed motors were sourced.
Currently the Baldor BSM100N-1250AA is looking to be the best option with a peak torque of 56Nm.
24NM motor Baldor BSM90N-1150AF
4m Boat Sail[]
For the 4m boat sail, a custom circuit was needed to allow an arduino to control a large (12V, 5A) motor. The current system employs the arduino controling npn BJT transistor that supply current to switch relays. There are 2 signals, one to supply power to the motor and one to control direction.
The circuit diagram for the sail motor.