Shop OBEX P1 Docs P2 Docs Learn Events
Need help, linerar actuator control via servo electronics — Parallax Forums

Need help, linerar actuator control via servo electronics

Aeroengineer1Aeroengineer1 Posts: 2
edited 2006-02-18 13:35 in Robotics
This seems like a place that might be able to answer my question.· I am going to be building a linear actuator using a servo motor and servo electronics.· There is just one complication, the number of rotations that this will have to do is about 160.· I need to build a small circuit that can replace the resistor and vary almost linearly the resistance from 0-5k over the total travel.· Due to space constraints, a linear variable resistor is not usable nor is it possible to get gearing to a 15 turn pot.· This needs to fit in a cylinder that is 1.5" in diameter and should not be more than 0.75" long.· I was thinking about using a hall effect sensor and a digital pic.· Any thoughts.· My electronics background is limited, but I can solder very well.· I am looking for ideas and information to understanding how to impliment these ideas.· Thanks in advance for your help.

Adam

PS this is for use with a RC submarine.

Comments

  • Bruce BatesBruce Bates Posts: 3,045
    edited 2006-02-16 06:42
    Adam -

    I'm not quite sure I understand the problem, but there are 3.5 x 360 degree rotation R/C servos (winch servos) as well as continuous rotation R/C servos available if that's of any assistance in this design. Winch servos can be found at places like Tower Hobbies, or your local hobby store.

    Continuous rotation and ordinary R/C servos are available right from Parallax:
    http://www.parallax.com/detail.asp?product_id=900-00008

    Regards,

    Bruce Bates

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    <!--StartFragment -->
  • Aeroengineer1Aeroengineer1 Posts: 2
    edited 2006-02-16 15:14
    Here is the setup.· Not only do I want to be able to drive the motor for the linear actuator, I want proportional control, ie I want to know where it is in its travel.· This is important for maintaining proper ballast.· So the problem arises due to the fact that I will be gearing the servo motor to an 0-80 thread rod.· To travel the approx 2" needed for blast it will need to turn 160 turns.· Because the standard servo pot will allow for only one turn and other 5K replacement pots go to 15 turns, these would require too much space to get the proper gear ratio.· Hence the need for a variable resistance circuit.· This circuit would replace the mechanical pot and feedback the position to the servo as if it were the pot it replaced.· A linear pot might work, but takes up a lot of space.· Does this help to understand?· Thanks for the reply.



    Adam
  • kelvin jameskelvin james Posts: 531
    edited 2006-02-17 07:00
    I take that you want to make a linear actuator that has some type of positional feedback, so you know where it is. There are a few ways to go about this. One is to use a stepper motor, it does not have feedback, but it's positional accuracy is very good if kept within it's torque rating. Small steppers can be had that could fit in your size constaints. For a continuous servo or small dc motor, an optical encoder for position, being rotational or linear, could work for you too. But it gets a little more complicated to read an encoder when trying to do other things at the same time, and may be overkill for what you want. The stepper would be the easiest / cheapest way , the encoder would be the fail-safe way for positioning.

    kelvin
  • bobledouxbobledoux Posts: 187
    edited 2006-02-17 15:10
    You might put a limit switch at one end of travel. Turn to this stop. Then use a hall sensor to sense each rotation of the screw, as it moves away from the stop. The microcontroller could count the turns. The microswitch would ensure "re-sync" against lost hall sensor readings. This approach probably requires interrupt capabilities in your controller, unless you are willing to stop other actions while you count the turns.

    I'd probably want the stop switch for re-sync even if I used a stepper motor. The question is whether lost steps cause a problem with your process.

    Suppose your stepper has 72 steps per rotation. For 160 screw turns you have 11,520 steps. That's a lot of steps to count.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2006-02-17 16:48
    If the PIC and hall-effect is too far fetched, I recently·mentioned to someone else that inside the 3 1/2 floppy drives are some very nice little stepper motors. If they can adequately drive your mechanical setup, I think they would be smaller than the servos and provide a cleaner approach to digitally generating your 'curve'.

    Just cannibalize a couple of floppy drives, have a ULN2308 drive the stepper, and go.
    The stop/re-sync switch remains an excellent addition.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)

    ······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

    Post Edited (Kramer) : 2/17/2006 4:51:22 PM GMT
  • gibbmangibbman Posts: 98
    edited 2006-02-18 00:08
    To Kramer's point, the floppy drive stepper might be a good choice. I noticed you mentioned 2" of travel, I don't know if the 3.5" FD will go that far, but if you have an old 5.25" drive, it might be plenty.
    Jim

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    In the end, it seems that it's all about getting the LEDs to blink....
  • ktekxktekx Posts: 71
    edited 2006-02-18 10:29
    How do you determine how many steps a stepper motor would require to take to do a full revolution? Is there a standard number or does it say on the stepper itself?
  • Bruce BatesBruce Bates Posts: 3,045
    edited 2006-02-18 10:52
    ktekx -

    The key parameter here is step angle which will be a parameter supplied with all stepper motors, either on the nameplate or in the documentation. 360 / step angle = steps per one revolution. Thus, if the step angle is 1.8 degrees, which is not uncommon, you have the following: 360 / 1.8 = 200 steps per revolution. Simple as that.

    Regards,

    Bruce Bates

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    <!--StartFragment -->
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2006-02-18 13:35
    For some odd reason [noparse][[/noparse]seems very strange to me at least] steppers usually come in 100, 200, and 400 steps per 360 degrees of rotation.
    200 steps seems to be the most common.

    If you start in on the math, it will begin to reveal how this is a 'neither fish nor fowl' kind of system. Traditionally everyone does angle calculations in 360 degrees, but if you read the BasicStamp Manual you find there is a Binary system that seems to rely on 256 portions of a circle.

    Fortunally 200 is less than a byte, so that is likely to be most convienent. You may have to cannibalize the motor and just test it for resolution as these are installed. Additionally, you will have to figure the coil resistance and determine if it is 5 volts or 12 volts. When in doubt, use 5 volts.

    The 3.5 drive might barely reach 2", but since 3.5 / 2 is 1.75" it is more likely a near miss. It you can just use the motor and attach a shaft, it is a very nice little stepper. I don't think it is easy to buy one this small.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)

    ······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
Sign In or Register to comment.