Droidbot 07 - Wheeled-Quadruped

Testing a simpler (pitch-roll) mechanism with a really nasty thumbstick. A 1L water bottle is used to see how it handles some weight, stuck on with one of those sticky phone pads from Banggood.

Next step; add an IMU for self-leveling. May need new (bigger/stronger) 3D printed parts.

OK, I'm admitting defeat with the Stewart platform idea. My 3D printed "heavy duty" servo arms aren't up to it (slipping on the servo shafts) and I'm spending more time on it than I wanted; to the point where it's blocking me from moving on and it's very annoying. I need to have a think, simplify the mechanism and get better servo arms.

Below is a brief video where you can just about see how the servo arms slip and that I haven't got the tilt rotating about the correct point. Crap. The shell of the head assembly would have sat over the top platform and flush with the body when in its lowest position.

It's been a while so I thought I'd post a quick something...

Mounts and arms available on Thingiverse.

A while ago I created some parts for the heavy duty standard sized servos that I'm planning to use in the bot's "head" mechanism. They nothing fancy, but they're available on Thingiverse:

1. Vertical servo mount.
2. 50mm heavy-duty servo arm.

I have six servos arranged around the head to provide the pitch and roll that I want (to keep head/sensors level as the body moves around) in addition to be able to raise and lower a little bit. It is effectively a Stewart Platform but I don't need yaw or horizontal translations so it should make the maths a little bit easier. The head spin motor and index sensors are function but the control code could do with some further love.

Still early days.

I have also been thinking about the legs, specifically how to mount and actuate the wheel/foot/ankle. I'm thinking of using the BLDC hub motors of "hoverboards". Though the name irritates me their motors/wheels seem well suited to my requirements. I modelled how I could actuate a foot but quickly realised it'll need more thought. The flawed concept is shown below.

Flawed concept.

Also, I'm currently printing an enclosure for a development remote; a battery powered platform with a 20x4 LCD and 3-axis joystick and some buttons, linked via Bluetooth. Intended as a tool to help early debug and give me a flexible terminal for simple display and control. Had to split the main body in order to fit it on the build platform of my printer.

W.I.P.

After two failed prints at 30% (corrupt file on a failing SD card it seems), I got my very own custom motor mount! It's my first custom print (I used Fusion 360) with custom slicer settings and was uncertain of the outcome. I'd previously only printed a Thingiverse Venus Box with stock Cura settings as a familiarisation exercise.

I need to work out the other half of the mount, pivoting in the bottom corner, but it's a step in the right direction.

PLA, 0.2mm layer height, 0.3mm nozzle.

25mmØ motor and 22mmØ bearing, M8 bolt as shaft, M3 nuts and bolts for clamping. Rubber wheel is from my junk collection, pulled from an old printer.

I want to use ABS in the future (so I can solvent weld parts) but have read that ABS can give off nasty fumes and as my printer is at work (design at night, print in the day and I can keep eye on it) and I don't want to gas the whole department, I should build a fume hood first. The hood with also deaden some of the noise too.

If anyone is interested in the model, I can post the files.

It took a while, thanks couriers, but it's finally here! Contents checked and assembly started.

Once built and tuned (hopefully not too long) I should be able to move forward with the head mechanism and try-out a couple of drive methods that I have in mind for the legs.