Shop OBEX P1 Docs P2 Docs Learn Events
$3.79 Digital Servo — Parallax Forums

$3.79 Digital Servo

ercoerco Posts: 20,255
edited 2014-06-22 14:28 in Robotics
First time I've seen this from Hobby King, Turnigy's "new standard" (sic) TD9D digital 9-gram servo. Possibly the cheapest digital servo around, and if they are as robust as HK's beloved HXT900, this could be great news for bot builders. http://www.hobbyking.com/hobbyking/store/__56272__Turnigy_TG9d_9g_1_8kg_0_09sec_Digital_Micro_Servo.html

Futaba PDF on digital vs analog servos at http://www.futaba-rc.com/servos/digitalservos.pdf

Duane, Rich and anyone with extensive servo experience, please weigh in on these. I have only used digital servos with standard analog servo signals. What are the immediate advantages, and is there a Propeller object which outputs digital servo signals?

Comments

  • Duane DegnDuane Degn Posts: 10,588
    edited 2014-06-20 10:52
    erco wrote: »
    I have only used digital servos with standard analog servo signals. What are the immediate advantages, and is there a Propeller object which outputs digital servo signals?

    These would also use the standard 1ms to 2ms pulse for position control. The difference with digital servos is some can receive (and act on) these pulses faster than 50Hz. I think most digital servos don't make use of the faster refresh rates. I know some helicopter tail servos can use faster refresh rates and many helicopter gyros have a switch to set depending on what kind of servo is being used.

    The fastest possible refresh rate is close to 500Hz but I doubt there any many (if any) servos who would use this sort of rate. I haven't ever measured the rate between my nice gyro and servos. My guess is they use somewhere between 100Hz and 250Hz rates.

    I learned from Jason Dorie that many ESCs can take advantage of higher refresh rates.

    The standard Servo32v9.spin servo driver which can control 32 servos with a single cog uses a 50Hz refresh rate. I'm pretty sure Jason Dorie has modified the code to use fewer servos (or ESCs) but at a faster rate.

    The main advantage using digital servo would provide to robotics projects is having the servos accurately hold their positions. As I learned from the pdf you linked to, analog servos provide current to their motors proportional to the difference between the present location of the servo and target location of the servo. This means if the servo is just a little off target, it won't try hard to make the small correction. A digital servo will use full power to move the small distance if required. This comes at the cost of higher current draw overall for digital servos.

    Some digital servos can have ramping values and end points set through the signal line. I've never used these sorts of servos myself. I've use Dynamixel "servos" which have all sorts of features but these don't use the standard pulse length for position control and are a completely different beast. There are servos which use pulse length to set their position but have additional features which can be accessed with proprietary programmers (these are the sorts I haven't tried).

    I've sure I'll get a few of these new servos with my next Hobbty King order, but I'll be surprised if I like them more than the HXT900 servos.

    One of the (many) things I like about the HXT900 is it has a full 180 degrees of motion (you have to extend the pulse range to .5ms and 2.5ms). I'm very curious what sort of motion range these new servos have.
  • W9GFOW9GFO Posts: 4,010
    edited 2014-06-20 11:52
    In addition to being power hungry, digital servos tend to buzz unless completely unloaded. They may also be less forgiving of imperfect pulse timing - although I have found some analog servos are also picky.

    Analog servos have a spongy feel to them when you press on the output arm, as Duane explained they don't try very hard to maintain position. I am speculating that this sponginess can actually be an advantage depending upon the application for protecting the gear train, such as in a hexapod. The gears will be under more stress in a digital servo. Lastly, you will still be at the mercy of the cheap pots that they use for positional accuracy.
  • W9GFOW9GFO Posts: 4,010
    edited 2014-06-20 11:55
    erco wrote: »
    and if they are as robust as HK's beloved HXT900

    This taken right of the product page:

    Important.
    The TG9e does not have the same strength and longevity as the HXT900.
    Please ensure your control surfaces are bind-free and always check servos after a hard landing or crash!
  • ercoerco Posts: 20,255
    edited 2014-06-20 12:10
    I do see that on the TG9e page, perhaps that applies to this new digital TG9d as well.

    TG9e: http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=9549
  • W9GFOW9GFO Posts: 4,010
    edited 2014-06-20 13:24
    Oops, clicked on the wrong TG9. I think it is pretty reasonable to assume that the only difference is in the electronics, so the warning probably is applicable.
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2014-06-22 14:28
    Nah. These are "digital" only in that they are using cheaper all-digital control circuits. Less expensive to make than all-analog servos, which more and more aren't all-analog any more.

    A *digital* servo, as defined by the more expensive models from Futaba, Hitec, and others uses a combination of digital control circuitry, typically a beefed-up motor (possibly coreless), and then due to the high currents to drive these motors, some seriously robust H-bridge circuitry (usually inside the digital control chip itself).

    Any of us could make a "digital" servo by hacking a cheapo model from HK, programming a 70 cent PIC OTP that has an analog input for the potentiometer, adding some 30 cent transistors for the bridge, and call it done. This would make a digital servo in name, but not in spirit. The mechanics are still Smile.
Sign In or Register to comment.