Newbie question: Why do my servos only run half speed?

Keezinator

I·am having a question: Using "Servos for PE Kit.spin" 1.4, only speed values between -200 and 200 seem to have different speeds.

Below -200 or above +200 the servos seem to have the same (maximum) speed.

Is this 'normal' or am I doing something wrong? The servos object allows speeds from -1000 (.5 ms pulselength) to +1000 (2.5 ms pulse length).

I·am using a propstick USB with the propeller education kit on top of a BOE bot chassis (and using the BOE bot servos).

I don't have any other power consumers on the board and I am driving the servos from the 5V power supplied by the regulator on board, while the controls are connected directly to the propstick output pins.

The whole thing is powered by 4 AA batteries...

Attached a picture of my setup ...

Post Edited (Keezinator) : 8/27/2008 4:02:19 PM GMT

Mike Huselton

I vaguely recall that the clock frequency is different on the Propstick versus the Protoboard. Is that the case? You may have to alter the PLL value.

Somebody with a fuller understanding of the PropStick and Protoboard differences may help.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
JMH - Electronics: Engineer - Programming: Professional

Post Edited (Quantum) : 8/27/2008 4:06:35 PM GMT

Keezinator

hm, interesting thought. You think the pulse length is not 'as advertised'? I'll have a look at the code and pull out my scope to see.

what I remember is that the "Servos for PE Kit.spin" code defines the pulse length in terms of the clock frequency, so any frequency should work.

its is 80 MHz by the way ...

any idea what max speed (RPM) I shoud expect from those Servos?

Graham Stabler

Perhaps the servo has just reached it's maximum physical speed based on the load.

Graham

Keezinator

Thanks. I figured it out.

Max speed is about 60RPM at 6V.

So I bypassed the regulator and set the pulse length to MAX and established that the servos are running at about 60RPM.

So all is working fine.

Next I tried to establish at what pulse wide maximum speed was reached: 1.5 (neutral) + .4 ms = 1.9 ms (approximately) where the max pulse width supported by the servos object is 2.5ms. So it tops at about 40% between min and max pulse width...

Right! So different servos seem to reach max speed at different pulse width (apperantly)...

Anyone know a good URL explaining all this (in depth)?

Mike Green

You're probably using what are known as "continuous motion servos". These are modified to remove the mechanical stop and the feedback potentiometer of an ordinary servo (or built with them absent). R/C servos were never intended for accuracy or precise speed, so some variation in maximum speed and the relationship between speed and control pulse width is expected. The only well controlled thing in ordinary R/C servos is the reproducibility of the position vs. control pulse width, not the absolute position, but the reproducibility of the position.

For more information, start with the Wikipedia page on radio control (R/C) servos.

Keezinator

Thanks Mike.... and you are right, I am using a continuous motion servo...

Phil Pilgrim (PhiPi)

Take a look at the Parallax publication, Robotics with the Boe-Bot (v2.2). On page 115, there's a graph (reproduced below) that illustrates the transfer curve relating velocity to pulse width.



As you can see it's highly non-linear, with most of the change occurring around the middle of the curve.

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Still some PropSTICK Kit bare PCBs left!