Uncategorized

Building a 5DOF robot arm 3


Here’s the latest on the 5DOF arm.

Tomorrow I’m preparing the last Makelangelo kits that will ship before the Vancouver Mini Maker Faire. I spent the better part of the day connecting the forearm to the bicep based on yesterday’s design. My design software crashed and corrupted a few files. Boo! I also experimented with limit switch positioning. I think the two opposite sides of the wrist are best – I can catch three different kinds of dangerous movement with only two switches. Plus the wires all run the same length, which makes them easy to measure and bundle together. The next time I rebuild the arm I’ll have specially placed holes for flawless switching.

Am I starting to experience diminishing returns? The tiniest change can have a trickle down that affects 4 or 5 laser cut parts. I’m resisting the urge to scratch at parts of the design that “itch” but still work. I tell myself they’ll get better soon enough. Keep your eye on the prize.

I can see how the shoulder and the base should look in my mind’s eye. Just like the switches it will take a few iterations to work out the details. I’m a little excited by the idea that I might not have enough power to lift the arm. What if I have to print a whole gearbox? Can I do it? How long would it take? It seems like a judicious combination of bearings and gears, both of which I’m learning about quite a lot these days. Heh – I thought these were good minecraft blocks, but it turns out maybe I can put them together to make an even better block.

Uncategorized

Building a 5DOF robot arm 2

Here’s a shot of my workbench right now.

arm5 may 31 physical

I went through three different wrists before I got the measurements right. The bevel gears were brushing against the forearm or the mounting screws. I also reprinted the wrist gears with a smaller center hole so they can hold the 5mm bolt that runs across the center. I need to find a way to attach that bolt to the wrist. A nut, a lockwasher, and another nut? Double nuts? Seems to work on the Prusa 3D printer hobbed bolt.

arm5 may 31

In the next version the elbow joint will change to increase the range of motion. In the current version there’s little or no room to add braces across the bicep pieces. Braces are needed to keep the bicep from twisting along it’s major axis. I haven’t worked out how to drive power to the elbow rotation yet. I’d like to put the stepper motor down at the shoulder again, but I might have to put it up near the elbow. Try it and see!

I’ve also started to plan the limit switches. In the Arm3 the switches are very close to the shoulder. There is less opportunity for error the further out they move. An error of 0.1mm near the point of rotation can turn into many mm out at the far end.

The total length of the forearm is something like 235mm. Once the tool changer goes on the end the distance from elbow to tool changer will be 250mm.

We got a mention in MAKE at the tail end of Robots Week. Thanks, Caleb! Please keep sharing with your friends. All my marketing is all done through word of mouth – usually my own. Your help drives sales, which gives me the resources to keep designing.

Uncategorized

Broadband to the Moon

According to Optical Society a new method has been developed to communicate with satellites, the ISS, and distant bodies like the Moon and Mars. Speeds are as good as broadband. Still no way around the 1.3 second delay tho, so don’t expect to be pwning noobz from Mare Nectaris any time soon. If I think about the per capita bandwidth I get a little too excited. This is a great day for science! Now we can all be space pirates.

read all about it.