Let's Make Robots!

Geared up Servo

Ok, thanx to Ignoblegnome idea I have done my 360° servo

First i started cutting pieces of PVC, i took a bunch of screw and nuts in order to build the frame for my superservo...

but suddenly Gareth appears in my mind and say: "Use the polymorph Luke..." :O

yeah right, very easy, 10 minutes of work and it's done, very sturdy and precice.

 

Now i can sweep my sensor in 360° :D

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Looking very smooth in the 2nd video. I hope it is meeting all your needs for this robot.

yep the head works fine

but it's hery hard to obtain a clean reading from srf08 :\

echoes are very unpredictable

i am thinking to trash it and to use a sharp sensor

Well, if you are having problems with the srf08, post a question and maybe someone here can help you figure it out.

there not much to say

in my experience this sensor work fine only to measure large flat surface perpendicular to sensor.

if you try to acquire data about small obstacles around the rover (a lighter on a table for example) it's a mess

sonic waves bouce in unpredictable manner and depends on materials and  edges..

for example an object with jagged edge may appear more far that its real position

another problem is with tilted flat surface that deflect wawes in another direction and not back to sensor...

 

furthermore the cone of sensig is too much broad and not simmetrical...

i have tried this sensor in a couple of robots, played a lot with Gain and Range register but with very unreliable results...

 

so i have to say: Sharp IR is better and this srf08 is crap in my humble opinion.

OK. I've never played with the srf08, but don't discount sonar in general. Many people on this site (including me) have had good luck with the srf05, which has better range than the Sharp IR sensors. I've also heard good things about he Parallax Ping sonar sensors.

I have one robot with a srf05 and another with a Sharp IR. Both sensors work acceptably for my purposes, and the Sharp is cheaper. ; j

I had a chance to test both US and IR sensors on the same type robot using the same code. The only difference in the code was the sensor reading. So here are my findings:

  • the Sharp IR sensor can detect much better the objects in front of the robot that are at an angle. The US wave gets deflected away and the robot needed bumpers.
  • unfortunately, the IR sensor has a lot of false readings when there is nothing in front if it
  • the US sensor (I use a SeeedStudio US sensor similar with Ping) has a better accuracy when measuring distance to large, perpendicular objects (walls)

So I think it is best to use both sensors on a scanning head. The IR can be used to draw a map of objects and the US can be used to check the position on the map by taking measurements to the walls or other known large objects. A compass can help the robot to stay perpendicular to the walls when measuring.

The genesis of this idea was a Shout Box discussion that led me to post this:

http://letsmakerobots.com/node/22879

Nifty, but I don't think I understand the application: Are you trying only to make a continuous rotation servo, or just one that does 360 degrees? Either way, isn't it more compact to just modify the servo to continuous rotation?

Gearing it up means you loose a bit of torque too - if that's a problem...?

or is it all about speed?

Continuous rotation is no good

because i have to take samples ad specific angles

metacode:

position servo to 0°

read sensor

position servo to 15°

read sensor

position servo to 45°

read sensor

and so on...

Great job! I'm glad my idea helped.

I may have to implement this some day.

Suggestion: Alter your post title and tags to include terms like '360'. It will make it much easier to search for.