Let's Make Robots!

mini robotic arm

moving a little bolt from one side to the other side

After having some ice cream I got the idea for this robotic arm.



Comment viewing options

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

Hey Steoerpeak!

Thank you so mutch for sharing this project and the idea with the teach in arm!

Inspired from your work i did it also with extra code for smooth movement:


Vielen Dank & merci villmool!

I've been pondering over a similar project for a few months now and was wondering whether the techniques used in assembling and programming this would be good practice for a larger multi-jointed robotic arm? I have no interest in making an automated system, just in having a robotic arm which will mimick the positioning of the user's control.

"Integrating the potis directly into the robotic arm is much more complex. You can use the potis inside the servos for that job they are still there. But you need to add an extra wire to each servo and you have to move the servo motors by hand while teaching in a sequence. And that is no good idea."

Why moving the servos by hand is no good idea? Due to the back currents that could be generated by the motors? Or is it due to the forcing of the servo motors when they are moved by hand?

Or are there others things that makes this a no good idea?





Hi Nick,

to move the motor itself by hand a little force is needed due to its permanent magnets, which create a small holding force. But inside the servo a lot of gears increase the force which you have to apply. If you move the servo by hand, you have to apply a much higher force. If you overcome the motors holding force, it starts to rotate an acts as a flywheel. So moving a servo by hand needs a high torque and its not easy to turn it to the position where you want it.

I think that the voltage which is generated by the motor is not so critical.?

( Fan control modules for engine cooling of real cars have some extra parts to clamp the voltage which is generated, when you drive at higher speeds. In that case the fan works like a windmill and creates higher voltages than normal inside the powerstage of the module.)

Greetings Stoerpeak


I would like to make this my 1st robot arm ever build.

Can you help me with the software and the diagram of the controller board?
What potentiometers did you use? Can they be integrated on the robotic arm so there is no need of additional arm for teach-in?


To rebuild this robotic arm,

you need to have some Icecream to get the basic material for the arm :-)

The Potis are standard types and are screwed to the white plastic parts with their nuts. The axles are pressed into the transparent plastic part. The handle to move the teach-in-arm is a M3-Spacer and the socket is a plastic part with is normally used to fix ballons on a stick to hold it.

The Software is simple: Its working like a servotester for four axis. That means, every poti is connected to an analog in of the controller and all servos are connected to GPIOs. The controller reads each poti, does some scaling, so that the angle of the poti equals to the angle of the axis and finally he creates the PWM output 1-2ms pulse every 20ms for all servos.

For "teach in", the controller has to save the values for each axis 5-10 times per second. So he can replay the sequence.

Thats it.

Integrating the potis directly into the robotic arm is much more complex. You can use the potis inside the servos for that job they are still there. But you need to add an extra wire to each servo and you have to move the servo motors by hand while teaching in a sequence. And that is no good idea.

The "teach-in" systems of industrial robots use force-sensors in each axis. So the teacher can move the gripper and the controller moves the arm (cotrolling the forces to zero). So the arm follows the hand of the teacher.

hi.. this project is great.. could you please help me with this?

Hi Stephen,

sorry for the late answer. but I was not here for a longer time.

And there were also a lot of other "problems" which had to be solved using my PIC Controllers.

i.e. self made electronics for a RC-Excavator which works similar to the digital system used in slot cars. To replace a lot of wires between the rotating part of the excavator and the track unit, a small PIC 12F629 reads the pulses from up to 5 channels of a RC-Receiver and leads their information via a 2 wire connection to a second PIC12F629. At the second board the power is seperated from the data. The PIC is reading the data, and generates the PWM output for the 5 Servo output connctors. the 2 wire connection is made with a cheap 6,3mm mono microphone plug which is also used as the axle for the rotating part.

The arm of the excavator is also powered by standard servos and so it was necessary to change the control behavior from proportional to integral so that the servos move like real hydraulic cylinders which are controlled by valves.

I used the same 12F629 type for that job and added some features like adjustable limit positions and starting point programmable by one jumper, and automatic return to park position when missing the pulses for some seconds.

But back to the mini robot. I will try to contact you by mail for more help with that project.

Now the clock shows 1.05am at my location (Germany). Time to leave


Greetings Stoerpeak


hi! can you send to me the source code of this project and the list of materials? i would like to make this as our project. thanks so much! my email add is stephen_baesa@yahoo.com.. tnx

I would like to make this soon.  It would be really nice if you could email me the list of materials to make the arm and the teach system.  Thanks, that would be great.  Email me at doggy1199@yahoo.com  Thanks