Has anyone tried this

Lynxmotion Robot Arms
1

I have been eyeballing these two items as a drive for the joints of an arm:

combined with this:

Curious if anyone has any experience with these, particularly the encoder and drives for the motors? Any info would be appreciated.

2

Hello,

here is a link from LM about making a monster servo. if you combine thos two servos, it’s just to powerfull, to much power, it’s pretty useless I think.

lynxmotion.com/images/html/ht01.htm

and secondly, the motors wouldn’t hold the position if you dont do like in the ‘‘how-to’’ and attach it to a servo. They will only do continuous rotations. but it would be interesting to have a monster servo on an arm, it could lift a LOT more. :laughing:

3

Encoders probably aren’t enough to determine precise positioning.
All you really need to make a servo, though is a speed controller, a microcontroller and a potentiometer (preferably a very linear one of at least 1/2 turn).

Stick the motor shaft onto the potentiometer’s shaft (using gears, pulleys, or whatever you’re comfortable with).
Wire the pot like so:

5V ------/\/\/\/\/\/\-------- GND
               |
               |
              Micro Analog Pin

And now you simply write a program that drives the motor towards the desired position.
The position will look like a value between 0V and 5V on your micro’s analog input.

If you want to get fancy and get a nice, smooth control, look into PID algorithms.
I’d suggest getting rudimentary control accomplished first, though, as you’ll need to understand that to do PID.

4

I was thinking that, but was just seeing if any one had played with the idea.
It would require a brake on the motor to hold position.

Encoders aren’t enough, it would require software and hardware to translate the pulses, but once translated the positional accuracy would be much greater.

I’m just really playing with ideas from what lynx has available in the store. I can see where this setup would be awesome on a rover for dead rekoning (have it fetch cold ones from the frig).

I was also looking into this as a way to move an arm in a programming envrioment that I am really good at. 3D laser programming in Gcode is for me really easy as it’s what I do every day. I’ve done it so long that I find that I actually think within the geometric limitations of each machine we have in our shop. With Rickety one I find myself continually attempting this type of thinking. I fall flat on my face as I want to think in an actual hard number for a point in space instead of the number required at the servo to be at that point.

It will just take me time. thanks guys

5

Encoders just emit a string of pulses that tell you the direction and speed of rotation.
You can easily write code to convert that to a relative rotation in degrees.

But, what it lacks is a set starting point.
If you turn your arm on, and the limb powers up with that joint bent at 30 degrees, it will think that the position is 0 degrees.

So, if you have a way to arrange the bot so that it always powers up with every joint in a standard position, then your idea would probably work.
Or, you could use a limit switch on each joint and have your robot “find” the switch.
Once it trips the switch, it knows that it’s at the angle that you mounted the switch on, and can take its future references from there.

You might get problems with buildup of error, the longer your encoders have been running, if you only have the starting position as a reference to go on, even assuming that the starting position is perfectly accurate.

You don’t need a physical brake on the motor to make it stop, provided that your micro is in control of the motor.
If your motor is going forwards, briefly telling it to go backwards in short bursts will slow it greatly.
Once it’s going slow, alternating between forwards and backwards very quickly will make it effectively come to a complete stop and hold a position.

This is a bad idea to try with crapy motors or motors that are heavily gear reduced.
Jim could tell you for sure whether or not the LM motors could hold up to it, but I’m going to venture a guess and say they they would.
All the motors that I’ve tried it on (which doesn’t include any LM motors) don’t seem to be bothered, with the exception of one el cheapo RC car.
If I recall correctly, I ended up stripping one of its plastic gears.