Small, Simple, Self-Centering Servo Circuit

Discussion in 'The Projects Forum' started by dafremen, Aug 18, 2011.

  1. dafremen

    Thread Starter New Member

    Aug 18, 2011
    14
    0
    Hello everyone,

    I've been working on converting a cheap fixed pitch coaxial RC helicopter into a variable pitch heli (for the challenge..not the savings..trust me..not the savings..:rolleyes:).

    So far so good..only the servo left to drive. Here's the setup:

    I'm taking the tail motor signal (3.7VDC LiPO powered) and want to use it to drive the Blade MSR servo, the smallest I could find:

    Figure A

    BladeMSRServo.jpg

    One problem I'll run into is that trying to get the servo back to hover will be a real pain without some sort of self-centering strategy.

    At first I thought, "No problem..I'll just have forward and use reverse to recenter back to hover." But that's Plan B. I really want Forward Backward AND Hover (because I'm greedy like that.)

    If you'll notice on the underside of the servo(Fig. A), there are brushes used for sensing position (is my guess.) I thought it might be possible to use those as a switch to charge a capacitor when the tailmotor signal is firing..then discharge the capacitor into the motor after the tail motor signal stops to bring it back to center.

    Here's the simple circuit I drew up:

    Figure B

    BladeMSRDIYServoCircuit.jpg

    I'm really not good enough with electronic design to see what sort of problems I've left for myself with this circuit. I also have no idea what size non-polarized capacitor it would take to actually drive the motor back to center. What sort of isolation do I need? I believe the tail motor circuit already has some isolation built in.

    Finally: Here is the circuit superimposed over the servo itself:

    BladeMSRServoTopView.jpg

    Hopefully you can see that the capacitor only charges and discharges when the servo is not centered. Charges when the tail motor is powered up..discharges when the tail motor is powered down. That's the idea any way.


    Any help making this circuit right would be greatly appreciated.

    Thanks,

    daf
     
  2. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,175
    397
    Do you have the resistave strips to make your unit a real servo? I do not hold much hope for a cap. drive. On a test to start a small DC [ 3V @ 40 mA ] motor at minimum current, required about 2000 μF @ 3V to turn it 15 revolutions. With a 3:1 reduction & a fine thread drive rod, .5 to 1 mm pitch, might take something like 20 to 50 motor revolutions to control pitch-- wild guess = 20,000 μF with slim chance of success. But this is not my field.
     
  3. dafremen

    Thread Starter New Member

    Aug 18, 2011
    14
    0
    Thanks for the quick response.

    Yea it looks like the capaticance required is overly large for the application.

    MY other thought was to have an "always on" circuit to return the motor to center that only turns off if the servo is centered, or if the tail motor is powered up.

    I have NO idea what it would look like.

    I'll take any help, suggestions or dictation I can get.

    Thanks again,

    daf
     
  4. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,175
    397
    An after thought: Motor ran 10 sec. from 3V charge on .2F @ 5v cap, 3mm X 11mm dia. The Electronic Goldmine, G17623 @ $ .99 on sale.
     
    Last edited: Aug 18, 2011
  5. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    What does the underside of the lid look like? The part where the fingers rub against?

    Great photo, by the way.
     
  6. dafremen

    Thread Starter New Member

    Aug 18, 2011
    14
    0
    Sgt. Wookie:

    There is no underside as of yet. That's what I'm trying to figure out here. In the diagram, The large thick black lines represent contact strips for the "fingers"/brushes to make contact with.

    Here is the challenge in a nutshell:

    There is a current which drives a tail motor. The tailmotor can be driven in either direction.

    I would like to take that current and use it to drive a servo motor instead.

    The only problem is that I have to bring the servo to center after the Tail Motor current is off again.

    If the tailmotor current drives the servo down, then I need a circuit to drive that same motor until the servo is centered.

    If the tailmotor drives the servo up, then I need that same circuit to drive the servo back down to center.

    Ideally, at no time should the centering function interfere with the tailmotor current driving the servo motor.

    The gear and motor set pictured above is the one I'll be using. It has those brushes on the underside. I'm assuming any solution would use those brushes to find center. It is a servo for the Blade MSR RC Helicopter.

    Thanks again for the help folks. This has been driving me a bit nuts, I'm afraid.. so every bit of help is appreciated. You Marine Corps Wookie? Was Amphibious Navy myself..back in the late 80s-early 90s. :D

    daf
     
  7. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,175
    397
    What is the nature of the tail motor current, reverse polarity or PWM & off??
    Add a back cover with a full length copper strip & a segmanted strip to denote max + travel, neutral, & minimum travel.
    Do you really want to always climb on right turns, sink on left turns, & neutral on no power?
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    After searching around a bit for "Blade MSR servo", found some relevant links:
    Problem with Blade MSR Servo:
    http://www.rcgroups.com/forums/showthread.php?t=1173034
    Blade MCX Servo repairs, including images:
    http://www.rcgroups.com/forums/showthread.php?t=1168805

    5-in-1 control board with servo partially removed:
    [​IMG]

    So, there are indeed a couple of strips under the servo to sense the position via the wiper contacts.

    The 5-in-1 control boards cost around $60 USD back in January of 2010, and glancing through the threads it seems that these boards are not without problems.

    I've forgotten much of what I knew about helos; only worked on them a little bit and never on the blade control systems. However, the main rotor blade angle is controlled by something called a "swash plate" assembly, which is controlled by the cyclic and the collective controls. What's really interesting about the rotor angle control is that any control inputs to the angle of the rotor blade are delayed by 90° before they have an effect on the aircraft due to gyroscopic precession.

    The mechanics of the swash plate connection to the individual blades is rather complex; the swash plate horizontal attitude can be at just about any reasonable angle off the perpendicular (Z) axis (tilting the swash plate) via the cyclic control input. For a very rapid version of the horizontal attitude angle change, have a look at this Euler's disk video at around the 30 second to 1 minute mark:
    http://www.youtube.com/watch?v=ug2bKCG4gZY

    Here is a high-speed video of a Euler's disk which more closely illustrates the attitude changes at a reasonable speed:
    http://www.youtube.com/watch?v=zpCJ_4Dg2zM&feature=related

    Along with that, the entire swash plate can be raised or lowered to increase or decrease overall lift via the collective.

    I can't see how a swash plate could be effectively controlled by just a single servo. The swash plate attitude angle changes enable the helo to travel in any direction or hover, while the raising/lowering of the swash plate changes the amount of lift.
     
    Last edited: Aug 20, 2011
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Found this video that illustrates the basic functioning of a cyclic control in a mockup:
    http://www.youtube.com/watch?v=vzbHiDbxhp4

    The collective control is not shown, as that would have added a considerable amount of complexity.

    One way to get around that complexity in a model might be to change the RPM of the main rotor. I can't think of an easier way to accomplish that.
     
  10. dafremen

    Thread Starter New Member

    Aug 18, 2011
    14
    0
    Thanks for the thorough reply Sarge.

    I'll be locking the sides of the swatchplate at hover position so that side to side won't need a servo..just frontward and backward. Since it's a coaxial heli, turn is controlled by decreasing the rate of spin of one of the main motors. Thrust is increased or decreased with a separate throttle contol.

    The crux of my problem is making this servo self center. Getting it into the up position is simple as is getting it into the down position. THat's just a matter of hooking the tail motor wires to the servo motor wires. However, without self centering, it'll be damned near impossible to make this heli fly in a stable manner.

    Got another idea..but no time to outline it..I'll be back tonight. Thanks again.

    daf
     
  11. Kingsparks

    Member

    May 17, 2011
    118
    5
    Dafremen.

    I have played a bit with coaxial hells, not much but some. My question is why do you want to center the tail control or motor separately when the transmitter (usually) does that? Centered as you say will result in a hover but your transmitter should center the motor control through the escapement.

    In other words, the one I have is much the same as any RC transmitter and selecting forward, (nose down or elevator down on a regular RC unit) drives the tail up allowing the helli to go forward. Back by moving the stick the other way.

    The transmitter I have, and most regular if not all, that stick centers by it's self allowing the transmitter to center the control. (Servo on a regular aircraft but escapement on the electric type hells). Also has a trim to fine tune the hover.

    I don't thin the capacitor idea will work because as far as I know all todays servos operate on a pulse train. :rolleyes:
     
  12. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Hmm - keep in mind what I mentioned about gyroscopic precession; the inputs you make to the blade angle will be off by 90° when the force is applied to the helo itself. This is difficult to explain in words, but this video of a bicycle wheel will help to explain what I'm talking about:
    http://www.youtube.com/watch?v=lvPAjr_a6Jg

    If the wheel is not spinning and one of the axle supports is removed, the unsupported side drops. When the wheel is spinning, the gravitational force that would have caused the unsupported side of the axle to drop is now delayed by 90°, causing the wheel to remain vertical, yet precess around the axle support point.

    I just don't want you to be overly surprised when the control output is delayed by 90°. It's not as intuitive as one would think. You may already know about this precession, but many others probably don't - so if you did, just consider this part as a possibly interesting tidbit about helicopter rotor dynamics for other forum readers.

    Ahh, I see. You did mention coaxial in your first post, but I had the idea you were going to a single rotor with blade control. I can only imagine that getting the blade angle adjustment of a pair of coaxial counter-rotating blades is going to be dicey at best.

    What voltage do you have to work with?

    A sliding potentiometer can provide feedback as to the absolute position of the servo; what I don't have a good idea on is the distance that the servo can travel. A sliding pot would have to have at least that much travel, preferably not much more.

    The signal from the wiper of that pot could be used as a voltage reference for a pair of power operational amplifiers, indicating what the actual position of the servo is.

    Here's what I'm talking about:

    [​IMG]

    The power opamps would compare that signal to a control input that is a simple voltage level. If the voltage from the position pot is higher or lower than the control input voltage, the opamps drive the motor in the direction to make the voltages equal by controlling the direction of current through the servo motor. Once the control input voltage equals the voltage from the position pot, the output voltage difference from the opamps is zero, so the motor stops running.

    This is not terribly difficult to implement, and requires only a few parts. But, I am not entirely certain how the control you have for the other servo works. It would help quite a bit if you could get some voltage readings.

    [eta]
    Just so you don't think that I'm overly clever, the circuit I posted above is very commonly used in automotive applications to control the position of the heater/AC blend door, to maintain the temperature of the cabin.
     
    Last edited: Aug 20, 2011
  13. Kingsparks

    Member

    May 17, 2011
    118
    5
     
  14. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I'm not certain on the configuration either. In the first post, our OP (original poster) said that he was converting a coaxial rotor assembly from fixed-pitch to variable pitch; I (incorrectly) assumed that they were going to a single rotor configuration; but I was corrected in a later post, which caused me to reply as I did. If the blade angles even momentarily happen to angle the blades to cause them to collide, the result will be a rather spectacular exodus of blades towards all points of the compass. :eek: I would not want to be anywhere nearby when that happened.

    Thanks for the confirmation. I was originally slated to be a helo mech, but they needed avionics/radar systems techs more than they needed helo mechs, so I changed my MOS from what I was guaranteed on enlistment.
    Thanks kindly for the compliment. :) I was never around helos while on active duty, but I've had some exposure to them over the last 8 years or so via a Vietnam War museum here in Orlando (we have a UH-1 aka Huey rigged as a medevac ship), and working on a friends' Brantley B-2B helo. I certainly don't proclaim myself to be some kind of expert on helos; just have a general idea of how they thrash the air into submission. ;)
     
  15. Kingsparks

    Member

    May 17, 2011
    118
    5
     
  16. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I had to think about it some more, and it finally hit me like a wet fish across the face...

    Counter-rotating blades, combined with the 90° gyroscopic precession, will result in the blade force's output being 180° out of phase. :eek: Even if they are aligned precisely in a static manner, when they are rotating, the blade thrusts will be offset 90° in either direction.

    As Hans Solo said many years ago; "I have a bad feeling about this..."
     
  17. Kingsparks

    Member

    May 17, 2011
    118
    5

    Are you trying to say we are going no where fast?
    Not a wet fish, just a #^%@$*&, why didn't I think of that and a big laugh. I still don't think I want to be close on the first test. :D
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Well, I may be wrong about my concerns - I'll have to think about it some more. It's making my head hurt, actually... ;)
     
  19. Kingsparks

    Member

    May 17, 2011
    118
    5
    Think about this.

    If you set one set of rotors for forward flight and the others for reverse + 90 degrees, wouldn't both forces work in one direction?

    Setting that up on an RC transmitter or receiver might be a problem. Or maybe you could connect the servos in reverse or just mechanically positioned to work out that way.

    Yeah, I'm fast catching that headache also. :eek:
     
  20. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Decided I'd try a Google search on "helicopter precession" and wound up with a bunch of hits, including this tutorial:
    http://helicopterflight.net/gyroscopic P.htm
    which explains single-rotor precession better than I did, and brings up a number of control issues during flight to be considered.
     
Loading...