Servo Speed Problem

Richard BaloghRichard Balogh Posts: 8
edited 2004-11-20 - 14:09:03 in Robotics
Hello,

my BoeBot has a combination of two servos:

·one is labeled STD (but actually IS modified) and is wworking OK:
· ·PULSOUT 750 = no movement,
·· 650 or 850 gives cca +/- 32 rotations per minute.

·second one is labeled PM (and also IS modified) and is working, but SLOW:
···PULSOUT 750 = no movement,
·· 650 or 850 gives only 12 - 16 rpm.

With this behaviour it is VERY difficult to achieve stright line movement.
Second (PM) servo has a small trimer inside, for callibration of the
center position, but it does not affect the final speed. Is there any
other possibility to achieve approximately same speed of both servos?

BTW: what is the actual schematic diagram of the circuitry inside
of· the servo? I did search the web but withou success.

Sincerely,
··············· Richard Balogh

Comments

  • glasoglaso Posts: 38
    edited 2004-11-05 - 12:06:45
    I had the same problem and I solve it using wheel encoders. With the encoders I can monitor which servo is runing faster or slower and correct it.
  • BeanBean Posts: 8,085
    edited 2004-11-05 - 14:06:48
    What if you use values lower than 650 or higher than 850 ?

    Bean.
  • edited 2004-11-05 - 20:29:26
    The PM servo should be turning in the 32 to 36 RPM range. If your PM servo has always turned that slow (12 to 16 RPM), it may have a manufacturing defect. If this is the case, please contact me offline, and I'll see if we can arrange a replacement for you.

    You can deliver pulses that slow down the modified STD servo to match the speed, but your pulses will probably have to be, in the 750 +/- (5 to 20) range.
  • Ken GraceyKen Gracey Posts: 6,607
    edited 2004-11-05 - 21:55:09
    Richard/Andy,

    I think Richard has some very old hardware - if I'm not mistaken this is one of the Boe-Bots I sent to him in Slovakia several years ago. He might even be mixing new/old servos together, from different manufacturers or even with the pot vs. fixed resistors.

    He probably needs a couple of new servos.

    Richard, how old are these?

    Ken Gracey
  • Richard BaloghRichard Balogh Posts: 8
    edited 2004-11-20 - 14:09:03
    Solved!

    The problem was, that initially I did use Test option of an GUI Bot
    software for measuring servo rotation speed.
    I did think it will be easier to do it without programming
    (even very simple).

    The key to the solution was question about the speed when
    pulsout argument is outside of <650 - 850> limits. Because in the
    GUI Bot one can not set arguments outside this limits, I (finally)
    write 3 lines of code and servo was repaired! The problem is that
    GUI Bot sends pulse train with too large period (50 ms). It is
    interesting, that one of my servos (labeled STD) is less sensitive
    to changes in repetition interval.

    The same "error" can be simulated by changing the repetition interval
    from 20 to 50 ms. To verify, I used an oscilloscope to measure
    repetition interval -- GUI Bot clearly uses about 50 ms. It is OK
    for the purpose of the program - to callibrate servos, but don't use
    it for measuring of the servo characteristics!

    Thanks for help.

    Richard Balogh

    Links:

    1. Graph of measured servo speeds

    2. Program used for measurements.
Sign In or Register to comment.