I need a circuit for positioning a linear actuator

Discussion in 'The Projects Forum' started by mwcesal, Feb 27, 2008.

  1. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
    I need to use a linear actuator to stop in 4 specific positions based on which of 4 switch positions are selected. It has a built in potentiometer. Does anyone know of a simple circuit?
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    What kind of switch? A 4-pole rotary?

    A rotary switch could be used to select a voltage from across a network of six resistors. This could be used for the non-inverting input to an op amp as a "goal" voltage.

    The pot on the linear actuator can be used as the "current position" indicator voltage, supplying the inverting input to an op amp.

    The output of the op amp would reflect the direction and magnitude necessary for correction of the actuator's position.

    This is oversimplified, of course. There must be feedback from the output of the amp back to the input to control the gain. There must be provisions for detecting "out of range" conditions, should either the pot or the rotary switch fail, so that the actuator doesn't attempt to slave to an unattainable position. You make no mention of the power requirements of the actuator.
     
  3. scubasteve_911

    Senior Member

    Dec 27, 2007
    1,202
    1
    As Wookie implied, you need a control system of some sort to accomplish your goal. You need to specify your desired system dynamics, which will help you select an appropriate typology of controller (fuzzy, PI, PID). Simple is a relative term for something like this. It may be a simple implementation (few opamps), but you may need to do a bit of math / iterative tuning to get it right.

    Steve
     
  4. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    mwcesal,

    Spec's on the actuator: voltage? current? speed? pot resistance?

    This a classic "Servo" system and can be very simple or very complex, as steve and SgtWookie have indicated.

    By simple do you mean: cheap or easy or quick? Pick two! (old joke in the shop)

    P.S. I'm currently implementing the steering system on a Fisher-Price Power Wheels-turned-Mars Rover for our computer lab, using a linear actuator /w pot feedback. ;)

    Ken
     
  5. techroomt

    Senior Member

    May 19, 2004
    198
    1
    you don't mention if the 4 positions are the only drive requirements to the actuator. if they are, then you could have the 4 digital switches select the appropriate dc drive voltages (representing the actuator positions) from a dac to the servo input.
     
  6. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
    Thanks for the quick responses. The switch will likely by a rotary switch. The actuator is rated at 10 amps, 12 volts, .5"/sec. I have not purchased the actuator yet, but the spec sheet doesn't list the ohms of the potentiometer. It would need to be able to move from any of the 4 positions to any of the positions, 1-4, 4-2, 2-4, etc.

    I'm no EE, but I can build a circuit from a schematic and can do math. Simple? Less than $100 to build and no programming.
     
  7. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    Couple more questions.

    How long is the travel of the actuator (position 1 to position 4)?
    How accurate do you need the positioning?
    Does the spec sheet indicate the positioning accuracy of the pot?

    That was actually three questions wasn't it...sorry. ;)

    Ken
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I suggest that the actual value of the pot that comes with the actuator is not that important at the moment, as it will simply be used as a voltage divider.

    However, once received, it would be a good idea to measure it, and select a pair of resistors that each measure about 12% of the value of the pot. This gives a useful output range of about 80% of the voltage potential placed across the network. If a voltage below 10% or over 90% of the voltage potential across the network is detected (using a pair of comparators configured as a window detector) the servo drive could be inhibited until the voltage fell within the window. Without such a scheme, the actuator (and whatever is attached to it) would be driven at maximum speed to an end stop, with a high probability of causing damage.
     
  9. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    One more question, in line with SgtWookie's comment...Does the actuator have internal limit switches? I just purchased one with an internal position pot, and it automatically cuts the motor drive in one direction at each end of travel. And he's right the actual value of the pot is not important.
    Ok, another question...do you have a link the the spec sheet?

    Ken
     
  10. beenthere

    Retired Moderator

    Apr 20, 2004
    15,815
    282
    I used a linear actuator in a soil penetrometer. It was a 12 volt monster made by Saginaw for opening and closing bus doors. There was a gear train that let me put an optointerupter in place to count gear teeth. Positioning was accurate over the 1 meter extension to <1 mm. But you almost have to have a uP for that kind of control.
     
  11. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    For what it's worth, here's an option.

    I'm using a PICAXE 08M µC. They're cheap (~$3.50), easy to program (BASIC), free program editor, require minimum hardware to program (serial cable + 2 resistors), and fast enough for this system (Interpreted BASIC at 4MHz). I chose to make an H-bridge of two 25A/12vdc/78Ω/SPDT automotive relays (two for $10)......because. ;) The four pots allow for easy adjustment of each position. However, you could have a 2-bit binary (4 state) input switch and have the positions set in the program code.

    Ken
     
  12. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
  13. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
  14. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
    Ken,

    I haven't programmed Basic in years. If I understand this, the PICAXE is taking pins 5 or 7 high depending on which direction the actuator needs go based on comparing the voltage at a switch position to the voltage at the actuator pot? It seems like a simple if-then statement.

    Mark
     
  15. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    Mark,

    That's the same actuator I'm using, except mine is the 2" version. Your concept of the circuit operation in exactly correct. The Basic programming is pretty easy to pick up again. I discovered these uCs a while back and was up and running in no time. I'm not a programmer, and hadn't touched BASIC on 20 years. PICAXE has good online manuals, and a very active, knowledgable, and newbie friendly Forum http://www.picaxeforum.co.uk/, so help is just a posting away. :)

    The PICAXE 08M has 10 Bit ADCs, so the positioning is 1024/4" or about 0.004" resolution and allowing for a few bits of dead-band it should be accurate enough.

    One thing to think about, with that accuracy, is position overshoot. When both relays are OFF, the contacts short out the motor and have a breaking action. Depending on the speed, breaking, and dead-band, you may have to go to to a slightly more complex program that anticipates coasting. Or you could go to a solid state H-bridge output driver with pulse width modulation for added speed control,...and...and...and. OK, now I'm just looking for problems. ;)

    Ken

    Added comment: I just checked, and pot in my actuator is 10K.
     
  16. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    While you COULD go with a design using relays, you will likely find yourself replacing the relays more often than you would like. Every time you make or break a circuit under power, the contact points will arc and burn a little. That adds up after a while.

    Power MOSFETS under digital control would be a good way to go - or simply purchase a pre-made H-bridge and control it.
     
  17. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    SgtWookie,

    I agree with the purchase of an off-the-shelf H-bridge. (I just happen to have that circuit available.) If the repositioning is only infrequent, then the relays would be OK, but we don't know....Mark?

    Ken
     
  18. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
    The positions will only change a few dozen times a week, but I like the idea of the H-bridge; decisions, decisions . . . . I've seen several designs on the web, is there one that would work best?

    When you say that the motor will short out when both relays are off, will that draw significant power while the actuator is "idling" or does power consumption go to zero once the actuator has reach it's position?

    The spec sheets show no connections to the end-of-stroke limit switches, are they accessible or just an over-stroke failsafe?

    I think it's time to start bread boarding.

    Mark
     
  19. KMoffett

    AAC Fanatic!

    Dec 19, 2007
    2,576
    231
    The relay "shorting out" the motor, only means that both leads of the motor are connected to the +12v. This causes the spinning motor to act as a generator producing a voltage of the opposite polarity, running into a dead short, causing it to stop quickly. No current flows through the motor on idle.

    I think the relays looked like a "quick and dirty" solution because the 4-position rotary switch make it sound like you were not going to be making frequent position changes. Since you mentioned <$100 for the project, there a lot of off-the-shelf H-bridge circuit boards for <$50. The hobby robotics market has driven these. :) (Kind of like video games driving the computer hardware advances.)

    They say to fuse at 10A, but don't give any other current data. My actuator runs about 1A with a light load. My guess that a 5A H-Bridge, with a 10A peak, would work. As said before, an advantage of the H-bridge over the relays is that you could control the speed and slow the actuator down as it approached a stop, for more accurate position control. I'm not sure, with active breaking, that that is necessary...just don't know.

    This is the controller I'm using for mine:

    http://www.newmicros.com/index2.php...ore/order.cgi?form=prod_detail&part=NMIH-0050

    It's a 5A, with a peak of 6A, and has thermal and short circuit protection. I'm not using the PWM, only forward-break-reverse. This one requires two PWM inputs to work with speed and direction control, and the PICAXE 08M only has 3 pins that can do ADC-in and PWM-out...you would need 4...so perhaps a PIXAXE 14M.

    What is the load on the actuator?

    Ken
     
  20. mwcesal

    Thread Starter New Member

    Feb 27, 2008
    7
    0
    I was thinking of solid state so I could pot the circuit and not worry about where the controller was located, inside or outside. But with 5A loads, there's bound to be heatsinks so potting is out. I've built several projects using relays and haven't had any problems.

    I haven't built the hardware yet, but I'm estimating about 20-40 lbs of load.

    Mark

    Mark
     
Loading...