Let's Make Robots!

Tiny delta parallel robot

Grabbing objects, swinging them around

This is a delta parallel robot - aimed to be the smallest in the world. I aimed at a height of about 1cm, but during building I found that, with my limited-precision home tools, I could not achieve sufficient accuracy to reach that level of tiny-ness. Hence, its maximum reaching height is about 3cm.

A delta parallel robot of this type, used for quick pick-and-place applications, is usually built up of a base, upper arms, lower arms and the 'head', on which an actuator (arm, magnet, air suction) is mounted. For the base I used MDF; for ease of construction (and shed-availability :-) ). The upper arms are make of brass (availability in shed), steel needles (cut to 1cm) and tiny beads (from my wifes DIY jewelry box).


The arms are made of cupper bead-clamps (from an art/craft shop, used to attach things to beaded necklaces). Two soldered together form the arm that connects to the base, over the beads of the upper arms.

Parts of the lower arms

Mounted upper and lower arm

 The head is created from pieces of brass and an M2 bolt for attaching actuators. Finally, the robot arms are powered by microservos (5 euro) by www.conrad.nl . The full robot (without an actuator) is shown here:

Full robot

The robots movements are controlled by a Parallax Propeller RPM board - 5V is available, and the output pins deliver sufficient voltage for the signal lines.

Currently the software is written for a set of demo movements, see the first video - nothing intelligent yet. I'm working on a set of building blocks (objects) for the parallax propellor especially for delta robots; including calculation of conversion of x,y,z to servo-angles.

Now I'm stumbling on the actuator, to grab and release tiny objects. Magnetic? Suction? Actual gripper? I do not know yet how to make it tiny. Well, a tiny electromagnet is easy, but if I use a ferromagnetic core, I'm afraid the remanence will be too large to release small iron beads, for example. Will be testing this shortly. Any ideas?

And another step: making a user interface. Probably in LabWindows. How to input 3D movements for direct control? Any ideas?

Update may 2010:

Finally found some time to continue to the Gripper:

Gripper of smallest delta robot

The gripper is made by simply bending a small piece of packaging plastic into a 'U' shape, and drilling two small holes opposing eachother. It is operated by a nylon wire in an isolation 'hose', compare it to handbrakes on a bike. A fourth servo pulls on the nylon wire to close the gripper.

The second video demonstrates the gripper operation. It grabs an M2 nut, swings it around, and puts it back into place.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Can't help you with your questions but a truly shagadelic little creature you build there :) Never knew that a bit of broken jewelery could look so funny.

Maybe for the gripper you could make a magnetically activated claw, perhaps by modifying a small solenoid for clamping with a spring release? Trying to think of something small enough to be mounted onto the head bolt while not obstructing the linkages is quite tough.

Awesome little project by the way =)

Cool! WHo call's Guinness ?


I call Genius!


Wow, that's  cool!

I would also suggest an electro-magnet, or magnetically actuated gripper. If you've got any really old disk drives laying around you might be able to scavenge some nice lacquered wire to make your own winding


Hi All,

Thanks for the compliments and for the idea on an e-magnet actuated gripper. After thinking for a short while, some ideas start to form - but this will be a challenge :-) I would like to make it approx 5x5 mm. The problem at these small dimensions is that surface friction is becoming very significant w.r.t. other forces. I'll be in touch!