Questions on steppers vs. servos

Discussion in 'General Electronics Chat' started by wannaBinventor, Oct 12, 2010.

  1. wannaBinventor

    Thread Starter Active Member

    Apr 8, 2010
    I've never used steppers or servos, but I think I'll need one of the two for a little project I'm thinking about tinkering with.

    Ultimately, I'm going to need something that actuates a flap back and forth, probably only 90 degrees or so. There really only needs to be 2 positions - up and down.... much like a modle airplane might actuate its flaps or rudder.

    The catch is - it MUST be low power. I don't have a mAh draw in mind, but I need it to be very minimal. The actuation will only take place once every several minutes.

    I understand that a servo works off of pulse widths, so I'm wondering if its possible to pulse it to a certain PWM it to a certain position, then turn off power to it and hold position. Is this possible, or would the solenoid just return to center once I cut its power?

    Generally speaking, am I going to use more power pulsing the solenoid or stepping the stepper (assuming same force applied)?

    Would a DC motor with limit switches perhaps be better (considering power) since I'm not trying to obtain actual resolution, just one extreme or the other?

  2. SteveSterling

    New Member

    Oct 12, 2010
  3. shortbus

    AAC Fanatic!

    Sep 30, 2009
  4. DonQ

    Active Member

    May 6, 2009
    I think the 'servo' he is considering is an RC, remote control, type of servo. Although I agree that this is quite a distortion of what I consider a true servo, it is what many experimenters consider to be one.

    The RC-servo has medium holding power when the supply is turned off. The gear ratio between the motor and the output shaft is so high that it takes quite a bit to spin the motor by torquing the output shaft. Any solution with a high gear ration would help, with a worm-gear being among the best. The bigger problem with an RC-serve is that sometimes they 'glitch' when they power-down, so the final position may not be where you think.

    I don't think that stepper motors would be a very good option. They are power hogs, and if you use a stepper that uses little power, it will also have very little holding power. A high gear-ratio would help here also.

    Maybe some sort of spring-loaded, electrically released brake would work. This would only require power when you want the shaft to move, and when powered-down would engage the brake, giving you however much holding-power you need.
  5. shortbus

    AAC Fanatic!

    Sep 30, 2009
    DonQ- the stepper that I suggested is a linear stepper. Instead of a rotating shaft, it has a threaded shaft. The rotating part is a nut, built into the motor, where the shaft would be.

    Because the moving parts are threads, they develop very high load movement with little power. Also because they are threads, they don't move when switched off. They have actually more holding force than moving force. Makes them ideal for moving and holding loads.
  6. Dyslexicbloke

    Well-Known Member

    Sep 4, 2010
    If you go with an RC servo you could always take the guts out and drive it with an external circuit.
    I have one that is connected to a pwm driver and a position sensing circuit that reads the origional Pot.

    The litle system only uses a single quad opamp and a 4 transistors, in a bridge, to drive the motor. The position it is in, and stays in if the power is off, is controlled by a single resistor connected to the circuit, I have 4 positions that I switch in.

    My unit drives the choke linkage on my generator and a simple timer sounds an alarm if the unit attemps to drive for more than 2 secs, which would indicate a problem.
    Essentally ita a voltage controlled positioner, I havnt bothered scaling for 0-10 but it wouldnt be hard.

    Of course I could have built a circuit that generated pulses and left the servo unchanged but this was easier, required far less precision, and has the advantage that the pot output voltage serves as a remote position indicator.

  7. DonQ

    Active Member

    May 6, 2009
    Linear actuator, yes, I've used similar items often, but usually the variety driven by a 'real' servo motor, and an encoder or resolver feedback. High performance versions compared to the stepper variety.

    The stepper motor versions you suggest also includes the high gear-ratio I suggested for any option, and works on a similar principle as the worm-gear (except 'inside-out'). But I think it is probably overkill.

    From his description:
    I think a linear drive is probably beyond what this particular question is asking for, especially if the other option is some sort of RC-servo. I may be wrong, but this sounds like something small.
  8. awright

    AAC Fanatic!

    Jul 5, 2006
    You haven't told us anything about how much back force your device will apply to the activating mechanism or the inertia of your moving element. However, consider a bistable mechanical actuator driven by discharging a capacitor into a solenoid. The capacitor charges up between actuations and stops charging (drawing current) when it is charged. The capacitor is sized to positively activate the solenoid under worst-case conditions of temperature and supply voltage. A power mosfet or an SCR discharges the capacitor into the solenoid on command from a manual switch or a timer.

    The tricky part is coming up with the bistable actuating mechanism. Take a look at the manner in which bistable push on - push off switches work. The secret is simply clever slotting of the moving and fixed parts and the manner in which the actuating bent spring wire is guided into one or the other side of a "V" groove depending upon the last position of the moving element. You can see the same concept at work in some latching relays with a pulse open - pulse closed action.

    If you arrange your device so the solenoid plunger or the mechanism is spring-loaded so that in the rest position the actuator holds the last position and the solenoid simply overcomes the spring momentarily to switch positions, you automatically have a position holding device. Then you simply have to be sure that the spring is strong enough to force the position transfer and the pulse to the solenoid is strong enough to momentarily overcome the spring.

  9. John P

    AAC Fanatic!

    Oct 14, 2008
    Another idea would be something like the "Tortoise" made to operate the switches on model railroads:

    This device uses a geared-down motor that's so low-powered, it just runs into stops and stalls there, with no servo or limit switches needed. When it stalls, the current is 15mA from a 12V supply, or less if the voltage is reduced. But is it physically too big for the proposed application?
  10. wannaBinventor

    Thread Starter Active Member

    Apr 8, 2010
    Good thoughts guys. Thanks for the help.

    I think I'm kinda fond of the solenoid actuator with a spring/brake DonQ suggested.

    It would go like this

    Code ( (Unknown Language)):
    2. GND~~~~~W~I~R~E~W~R~A~P~~~~~~~~~~~~~
    3. ~ |||||-N-A-I-L--WITH-CATCH NOTCH|||||S()P()R()I()N()G~
    4. ~_________________________BRAKE/CATCH_____~
    5. ~~~~~~~~W~I~R~E~W~R~A~P~~~~~POS VOLT~~
    The above is probably an unclear and silly diagram, but perhaps it gets the idea accross.

    The spring would keep the nail it in the left position. I could energize the coil to push the nail/piston through the cylinder compressing the spring and making it catch on the brake catch notch. To release it, I could use a similar system for the brake catch -- something that, when energized, pulls it down momentarily for the piston to be released and pushed back to the left position with the spring.

    Are there already small, low power, low force, spring loaded solenoids like this available or do I need to rig something up?
    Last edited by a moderator: Oct 14, 2010
  11. Markd77

    Senior Member

    Sep 7, 2009
    I think the formatting has got messed up.
    If you use the code tags it won't mess up the formatting and your text art will make more sense.
  12. bertus


    Apr 5, 2008

    I tried to clear the diagram with code tags, but it still looks strange.
    Perhaps drawing it would be better.

  13. thatoneguy

    AAC Fanatic!

    Feb 19, 2009

    I'm reading this as a solenoid with a "neutral middle". The problem with is that (as shown) it could only be pulled or pushed in one direction since one end of the coil is tied to ground.

    Driving it with a Relay in H-Bridge configuration would work to reverse polarity. Then the issue would be what material to make the "catches" out of so they will hold, yet not wear out or abrade quickly.

    This solution would also need high current pulses during switching, since a DC Solenoid electronically looks like a low value resistor.

    The "Glow Cubes" made by Daktronics used a unique approach to a simlar problem in the early 90's. These are the signs where the 1" square pixels literally "flip" between bright yellow/green and black (still seen in road construction signs). How it works is each pulse temporarily magnetizes a metal piece, which causes it to rotate away from the pole it is currently at. To reverse, it reverses the current through the coil. There is no holding current with Glow cubes. See patent by Daktronics for details.

    --ETA: Patent.. Patent
    Last edited: Oct 14, 2010