Shop OBEX P1 Docs P2 Docs Learn Events
New Digital 9G Continuous Rotation Servo — Parallax Forums

New Digital 9G Continuous Rotation Servo

ercoerco Posts: 20,248
in Robotics
The FT90R is nice and new from Feetech, and erco-approved from this quick benchtop test. Need to build a bot around these next.

Continuous rotation 9g servos are hard to come by. These are digital. Feetech also makes an analog CR servo but those didn't pass muster, as shown in the second video, which also shows some hacked HXT 900s.









Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    Those look nice. 'Can't find 'em online, though. Gotta link?

    Thanks,
    -Phil
  • ercoerco Posts: 20,248
    No sir, not yet. I'm the Beta tester!
  • MikeDYurMikeDYur Posts: 2,176
    Very quiet for MG, fast to.
  • xanaduxanadu Posts: 3,347
    Nice, I'd be interested in a couple when they're released :)
  • ercoerco Posts: 20,248
    edited 2016-12-14 01:37
    Not MG. Will have to try the digital control signals from my servo controller. That video used standard 50 hz analog signals.

    Edit: Controller had 125 and 250 hz options, which had no effect.
  • MikeDYurMikeDYur Posts: 2,176
    erco wrote: »
    Not MG.

    Sorry, missed it.
  • GordonMcCombGordonMcComb Posts: 3,366
    erco wrote: »
    Edit: Controller had 125 and 250 hz options, which had no effect.

    Only real effect would be increased torque. Maybe you could try some kind of simple yarn-pulling test.

  • ercoerco Posts: 20,248
    Now I'm confused. I knew that digital servos were backwards-compatible with 50-hz analog signals, but I though there was a new digital standard. I keep seeing references to 300 hz servo signals. I thought that was the new control signal input frequency, but after doing some more digging, it appears that digital servos still use the traditional 50 hz, 1-2 ms control signals, and the internal digital circuitry drives the motor at 300 hz PWM frequency.

    Who done it?
  • tonyp12tonyp12 Posts: 1,950
    "Digital servos are the future for model control, and anyone who has used them says the difference is so significant that they would never return to standard servos"
    http://www.futabarc.com/servos/digitalservos.pdf
  • ercoerco Posts: 20,248
    Continuing quips from that old but still-useful PDF:

    To quote turbine display pilot Steve Elias, ‘Digital servo response and precision
    is like flying on rails. After flying digital servos, analogue versions are like
    controlling custard’.

    In spite of those bold statements, I remain analog servos' biggest flan.
  • RS_JimRS_Jim Posts: 1,753
    Eric
    I am curious, The Feetech Motor controllers that you sold to me, are they a PWM digital out or strictly an analog voltage? My scope is not working correctly for me to check.
    Jim

  • ercoerco Posts: 20,248
    I have not checked on a scope either but I'm 99.9% sure it's PWM. It's very torquey and an analog DC voltage wouldn't do that.
  • RS_JimRS_Jim Posts: 1,753
    Gee, That would mean that you could use that along with some positioning sensor to create a "digital" servo.
    Jim
  • xanaduxanadu Posts: 3,347
    Erco is RS_Jim referring to the motor controller I have?
  • GordonMcCombGordonMcComb Posts: 3,366
    edited 2016-12-15 02:08
    erco wrote: »
    Now I'm confused. I knew that digital servos were backwards-compatible with 50-hz analog signals, but I though there was a new digital standard. I keep seeing references to 300 hz servo signals. I thought that was the new control signal input frequency, but after doing some more digging, it appears that digital servos still use the traditional 50 hz, 1-2 ms control signals, and the internal digital circuitry drives the motor at 300 hz PWM frequency.

    Who done it?

    There isn't just one type of digital servo. Some are OTP and you get what comes out of the box. Others are reprogrammable. Others are ____ -- fill in the blank. It really depends on the maker, and the cost points they're going for.

    I can see a less expensive digital servo not having the controller to "multiply" a 50Hz input to 300Hz, relying instead on an external circuit for it. This would be particularly so for a non-RC application, which fits the bill for CR servos. They're assuming you're going to use some kind of microcontroller where you can set the frame rate.

    I think it was Duane who experimented with sending some >50Hz pulses to an analog servo. It accepted those pulses to a certain frame rate, after which it got extremely glitchy -- possibly because the pulses were too short for the one-shot generated inside the servo that was used for the countering error signal. I recall his pointing out the servo got hotter with faster pulses, suggesting a higher power consumption of the motor. Don't think anyone ever attached it to a fish weight, but likely the higher current produced higher torque.

    There's also the issue that many analog servos produced these days are actually digital inside. Cheaper to make.

  • ercoerco Posts: 20,248
    xanadu wrote: »
    Erco is RS_Jim referring to the motor controller I have?

    Yep. Yours had a bad channel, right?

  • xanaduxanadu Posts: 3,347
    edited 2016-12-15 03:19
    I thought it was bad, but it's sleeping :)

    I wanted to post a video of the PWM output for RS_Jim. Although I couldn't get them to work on my FPV bot, both channels appear to be functional on my scope. I know they don't produce enough current to move a motor. It seems the voltage is there. Interesting.

    Here's the PWM output of the Feetech(?) motor controller.

  • RS_JimRS_Jim Posts: 1,753
    Xanadu,
    Thanks for the video!
    Jim
  • MikeDYurMikeDYur Posts: 2,176
    Now I can't see xanadu's videos, just white screen.
  • RS_JimRS_Jim Posts: 1,753
    Mike,
    I could not see them on my ancient Ipad, had to switch to my laptop with firefox.
    Jim
  • MikeDYurMikeDYur Posts: 2,176
    edited 2016-12-15 15:19
    Jim,

    My only choice is Samsung G S5 , this has happened to me before on other posters videos, it may be my network is not dealing with it the right way.

    I wonder who xanadu uploads his videos to.

    Publison helped me see the video the last time this happened.

    Mike
  • PublisonPublison Posts: 12,366
    edited 2016-12-15 15:30
    Try this YouTube link:


  • MikeDYurMikeDYur Posts: 2,176
    Perfect, thanks Jim.

    Nice little pocket scope xanadu.
  • GordonMcCombGordonMcComb Posts: 3,366
    Xanadu is using the "wrong" way to embed videos, whatever that is. They're links to Youtube, alright, but the code wrapping the link is different from the ones where you can actually play video.

    Since I have Flash disabled, I get a prompt to download the swf for the movie. I see the properly embedded videos, which play through an HTML5 video player.

    I think we need a sticky somewhere, written by Master Youtube poster Erco, on how to do it right. I know I can never remember the special incarnation, and have to try it a bazillion times before it shows correctly.
  • ercoerco Posts: 20,248
    I just paste the link in, without using the video button. IIRC that button causes some download to start.

    Thanks xanadu! Is it 50 hz PWM frequency driving the motor?
  • xanaduxanadu Posts: 3,347
    I noticed that if I don't use the tag you have to click the videos twice. I started using them on the joystick thread and that seems to have worked. Is it possible older video ID's don't work with the tag and new ones do?I'll go back to not using tags. At least that way if the video doesn't load the link should be there.

    My ailing scope says 15.5KS for frequency. I can try it on a known frequency to see if that is accurate. Before I send it to Seed Studio for repair, I will probably replace it with this (the next best cheapest thing) - https://www.picotech.com/oscilloscope/2000/picoscope-2000-overview
Sign In or Register to comment.