Shop OBEX P1 Docs P2 Docs Learn Events
Digital servo — Parallax Forums

Digital servo

deadeyedeadeye Posts: 2
edited 2007-11-21 20:57 in Robotics
I have two TowerPro MG995 servos that I would like to use with my SumoBot. These are described as digital, coreless servos and there is no potentiometer. Can anyone tell me if these can be modified for continuous-rotation, and if so, some info on the required procedure would be greatly appreciated.

Comments

  • Paul BakerPaul Baker Posts: 6,351
    edited 2006-01-20 13:08
    Hmmm, excellent question. Its not easy finding information on the servo, I don't think the company has a web presence (I think they are chinese, and few of them do). If you are correct that it does not contain a potentiometer and instead uses a direct from motor feedback mechanism to recieve position, then it may not be possible to convert it. Without being able to diasect the servo myself, I cannot say for certain.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    ·1+1=10
  • SN96SN96 Posts: 318
    edited 2006-01-20 14:11
    That is a good question. I could not find any info regarding digital servos converted to continuous rotation. I was only able to find sites that discuss the opposite, converting standard servos to digital. Since it has no pot, I would think the controller board inside the servo would need to be hacked.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike

    ·
  • Bruce BatesBruce Bates Posts: 3,045
    edited 2006-01-20 14:35
    Gents -

    Unless the prices of digital servos has dropped drastically, I wouldn't even consider hacking one. What's the advantage to hacking a digital servo over hacking an analog servo which is FAR less expensive?

    Regards,

    Bruce Bates

    Post Edited (Bruce Bates) : 1/20/2006 2:46:32 PM GMT
  • bulkheadbulkhead Posts: 405
    edited 2006-01-21 05:39
    Hmm, I think Bruce has a great point.

    Digital servos are digital to provide more positioning accuracy. If you are going to hack it for continuous rotation, it being digital doesn't increase performance at all. I guess since this isn't made in the US, and that I've never heard of this brand, it's probably an inexpensive digital servo, so it might not matter much (if you can figure out how to convert it).
  • kelvin jameskelvin james Posts: 531
    edited 2006-01-21 05:52
    After a little looking around, i think you may have the wrong info about the servo not having a pot. It seems the tower pro brand are basically a generic type , that other companies put there name on also, and is almost identical to a futaba. So, unless someone has done some type of modification to it, i would treat it as a normal type servo. That means the continuous rotation hack could be done the same way as on the analog servo( not sure about wiring though ), the hardware is all the same, it is only the electronic control that is different. There are lots of step by step tutorials around about hacking a servo, the easiest way to find out is to open it up and have a look, if you want to chance it. I added a nice little pdf that explains alot about a digital servo.

    kelvin
    pdf
    pdf
    digitalservos.pdf
    244K
  • deadeyedeadeye Posts: 2
    edited 2006-01-22 21:07
    Thanks for the replies, but so far no help.· I got theses servos on Ebay really cheap and I didn't think their being digital was going to provide an advantage... they just happened to have oodles of torque.· I also didn't think they would be any more difficult to hack for continuous-rotation than a regular servo, but after taking one of them apart, I'm afraid I was wrong.· My first clue that something was different was the fact that there is no hole to center the potentiometer on these servos.· An internal investigation only revealed some very nice metal gears, bearings, a circuit board, and a motor.· I understand there has to be some kind of feedback, but there definitely is no potentiometer and I can't figure out how the feedback is accomplished with this servo.
  • Bruce BatesBruce Bates Posts: 3,045
    edited 2006-01-22 22:42
    deadeye -

    This is purely a guess, but it wouldn't surprise me if they used a Hall Effect sensor. Not that that information is all that much help - sorry.

    Regards,

    Bruce Bates
  • Paul BakerPaul Baker Posts: 6,351
    edited 2006-01-23 17:31
    The newest wave of motor feedback is called Back-EMF which is used to drive brushless DC (BLDC) motors, but can also be adapted for use in a brushed DC motor driven by PWM. A motor can also be used as a generator (spin the shaft by hand and a voltage is produced on the terminals), it is this electricity being generated by the spinning shaft that is analyized during a period of the PWM being off that is used to measure the position of the rotor in relation to the stator (vaguely akin to an encoder). If this is the method being used on you servo, the feedback is internal to the controller and is not modifyable beyond replacing the controller.

    Like I said before, without having physical acess to the servo, I cannot determine the exact means of feeback used.

    Another method that is used as Bruce has indicated is hall effect sensors. Since the gears are metal, a biased hall effect sensor (a hall effect sensor with a permanent magnet attached to the non-sensing side) can detect the passing of the teeth of a ferrous gear since the presence or absence of a gear's tooth is detectable by the small change in the electric field caused by the gear tooth's transversal of the permanent magnet's field. Again this method uses an encoder type methodology and is not easily hackable.

    The best bet would be to investigate whether the controller supports reprogramming to alter it's stop values. Many digital servos support this but ussually require a dedicated programmer at an additional cost.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    ·1+1=10
  • SN96SN96 Posts: 318
    edited 2006-01-23 17:51
    I copied this off of another site that somewhat explains the differance between analog and digital servos.

    "An analog servo updates the motor every 20mS (0.02 Seconds). So if it drifts off the commanded position it will take up to 20mS to send the motor another burst of voltage. So the motor isn't used to it's full capability. A digital servo updates the motor every 400uS (0.0004 Seconds). So the motor gets these bursts of voltage much faster. The more often a motor receives voltage, the more work it can do."

    Digital servos provide better torqe and holding power.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike

    ·
  • Steve JoblinSteve Joblin Posts: 784
    edited 2006-01-23 17:55
    Can you strip out the electronics making it essentially a high-quality gear box? You could then easily drive it with an H-bridge or simple Motor Contoller like the Pololu Dual Serial Motor Controller.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2006-01-28 17:38
    Have you opened one up to see if there is a pot inside?

    At some point you need to look at the internal workings for feasiblity. Some servos actually have only a 180 degree arc of toothed gear.
    Even if you have a full 360 degree gear, there may be a plastic stop that requires removal.

    And if you could photo and post what you are looking at, the advice would be a lot less speculative.

    Are you interested in trading these for more conventional servos?· I suspect someone would be happy to have the digitals for their original use.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "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) : 1/28/2006 5:43:13 PM GMT
  • Robot JayRobot Jay Posts: 12
    edited 2006-12-01 21:23
    This thread is REALLY old, but I thought I'd throw in my experience with these servos, just for posterity's sake.

    I purchased 20 of these MG995 servos for use in my bipedal walker.· I opened the servo, and there IS a potentiometer, but these servos are REALLY cheap, and the potentiometer is equally as cheap.· They are NOT digital.· They are just standard, metal gear, servos.· They could easily be modified for continuous rotation.· In fact, the potentiometer is so shotty, that removing it altogether would improve the functionality of the servo.·

    Just my $.02.· Later.


    -Jay
  • SchiftySchifty Posts: 1
    edited 2007-08-22 03:34
    HI There Robot Jay,

    A little late, but here is a picture of the MG995 Tower Pro Servo.· As you can see, there is a stop on the gear on the left.· Removing it should allow for continuous rotation.



    Enjoy· Jeff
    MG995Servo.jpg
    320 x 240 - 51K
  • Joe "Bot" RedJoe "Bot" Red Posts: 81
    edited 2007-08-23 03:17
    the gear assembly looks quite similar to the Hobbico CS-80, I actually modified one for continous rotation... look here:

    http://forums.parallax.com/showthread.php?p=660797


    At the bottom of this thread you will find a link to the instructable of how I modified it..

    I hope this helps!..

    Happy Roboting!!!
  • TEIBTEIB Posts: 2
    edited 2007-11-21 19:45
    Deadeye,

    I am just curious, were you able to use the MG995 for continuous rotation? I noticed some on ebay for cheap (2 for $25) but they say they are digital, so I am not sure they can be easily hacked. Has anyone actually hacked MG995's for continuous rotation?

    Thanks,
    -Tracy
  • LilDiLilDi Posts: 229
    edited 2007-11-21 20:57
    I did some research on digital and analog servos a while back. What I learned was that digital servos are generally faster, more accurate and provide more torque than analog servos. Other than that, there mechanics are the same. Analog servos are nearly extinct. You would be hard pressed to find one anymore, with the exception of many RC airplane servos are still analog because they are lite weight
Sign In or Register to comment.