DPRG
DPRG List  



[DPRG] Servo question

Subject: [DPRG] Servo question
From: paradug paradug at gmail.com
Date: Tue Nov 22 14:09:12 CST 2011

John,
    I think Tom has hit it on the head. Essentially if the servo is 
underpowered as it reaches the control band that surrounds the dead band, 
the driving signal and power is proportional to the remaining difference 
between your setting and the actual location. If you are stopped, it will 
not get to the set point due to friction or other power requirements. If you 
are moving, you will get there on the next servo period, since the the 
friction is less. Digital servos generally can divide the 1 ms to 2 ms 
standard servo range into 1024 parts. Analog servos do not have that 
ability. As Rud explained hobby servos do not have a PID.

    Generally the more you pay, the better the servo. However I have had 
good luck with some of the larger EXI metal gear digital servos from China 
for robots. Depending on your needs they range from about 8 to 16 dollars. 
For small tasks, like rotating sensors, the little servos from China which 
sell for about 3 dollars are fine. My favorite standard size, non-demanding 
application servo is the 41G T-Pro SG5010 Ballraced Servo. It costs about 5 
dollars. I generally buy from Hobbypartz (they are currently out of stock on 
all most servos, which is very unusual). If you want a standard sized 
continuous rotation servo for a small robot, I suggest the SpringRC 
SM-S4303R Continuous Rotation Servo sold by Pololu for $12.95. They have an 
externally accessible zero setting and seem to have an wider than normal 
control band.

 If you would be interested in a group buy from Hobbypartz, let me know. 
They carry a lot more than just servos. Orders over $50 have free shipping.

Regards,
Doug P.




--------------------------------------------------
From: "Tom Brusehaver" <cozytom at gmail.com>
Sent: Tuesday, November 22, 2011 1:04 PM
To: "John Swindle" <swindle at compuserve.com>
Cc: <dprglist at dprg.org>
Subject: Re: [DPRG] Servo question

> I havn't seen your specific robot, or servos, but I'll offer
> some explaination.
>
> Most servos have a feedback loop. They will move to their
> position. If they are close, they won't go to full power, and
> will run at a lower power setting to prevent overshoot. Once
> they are at the proper position, the motor is "off". They may
> momentarily turn on to reset the position if the position
> sensor (usually a simple potentiometer) detects it is out
> of position.
>
> Moving a large distance works well for an undersized servo
> since it will use full power to move a long distance, since
> there is reduced risk of overshoot. Also once moving, it will
> be easier to stay moving, as it approaches the target position,
> since it is already over came inertia (sticktion, and other factors).
>
> Using a small servo risks tearing up gears. The gears are only
> rated for a little more than the servo is normally planning on
> dealing with (torque wise). If your servo is hit hard (either in a
> fall, or bumping into a wall) it could exceed the limit, and tear
> the teeth off the internal gears (they can be replaced).
>
> Using a larger servo may use less electricity since it will
> be in the proper position more, than using an under
> sized servo. The Larger servo will get to the position, and
> turn off much quicker.
>
> Another option is to change your arms. Using smaller arms
> on the servo (or holes closer to the center) will require the
> servo to work less since they will exert the same force over
> a shorter distance (less inches, same ounces = more ounce
> inch torque).
>
> Also check what you are moving. If you have a lot of friction
> in what you are moving with the servo, your servo is working
> to overcome that, instead of doing what you need it to do.
>
> Shop around, older Futaba S8 come up every once in a while
> for about $10 on ebay and such.
>
> On Tue, Nov 22, 2011 at 12:45 PM, John Swindle <swindle at compuserve.com> 
> wrote:
>> Folks:
>>
>> Before Bruiser, I'd never used servos, so I'm learning a lot that
>> y'all already know, and I'll ask some pretty simple questions.
>>
>> Do the common servos that cost $20 to $40 have an I-term, an
>> integral term in their internal controller? I ask because I don't
>> understand why Bruiser's steering servo won't always restore his
>> steered wheels to straight when the servo is given continuous
>> neutral 1500 microsecond commands. The wheels go straight if I
>> unload the weight (pick up the front), or if I apply throttle.
>> I'm told that I need more ounce-inches of force, but that's not
>> true because I can manually command the servo to move the wheels
>> to straight by over-compensating the command. If I can do it
>> manually, then why can't the servo's own integral term do the
>> same thing? What am I missing?
>>
>> And, more importantly, how can I force the steered wheels to
>> point where I want them to, without putting a $200 servo on the
>> platform (such a servo would deplete the battery charge rapidly)?
>> Are servos inappropriate for steering?
>>
>> Thanks,
>> John Swindle
>>
>> _______________________________________________
>> DPRGlist mailing list
>> DPRGlist at dprg.org
>> http://list.dprg.org/mailman/listinfo/dprglist
>>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at dprg.org
> http://list.dprg.org/mailman/listinfo/dprglist
> 

More information about the DPRG mailing list