logo

More kinematics and updates

I’ve been busy. I have now learned how to make “gif” images, so I made some gifs of the kinematics that I had been playing with. This helps to understand the maths a whole lot more, and give an impression of what is going on. Here is some of the first ideas:

Vertical_Linear_Parallel_System_Perspective
Horizontal_Linear_Parallel_Inverted_System_Perspective
Horizontal_Linear_Parallel_System_Perspective

 

Lately most of my time has been working on a better model of the above image, after all it is pretty crude. The maths has been sorted, but now I need a working model which can exist in the real world as the phase 2 prototype. It’s almost done. I will post an image shortly.

The modelling has been pretty full on, so as a break I took the time to try and drive the rotational rig via a wiimote. I thought it would look pretty cool and that the kids might like to have a direct way to move the rig, rather than punching in numbers to drive to a position. All in all it worked well enough. I wrote a script in python (my first) to take data from the wiimote using ‘cwiid’ (http://abstrakraft.org/cwiid) and porting it to the arduino via ‘pyserial’ (http://pyserial.sourceforge.net). I know it’s been done before, but this was my first attempt at a script and I wanted to see what was involved.

The end result was a bit underwhelming. I think it’s a combination of my arduino sketch (which drives the rotational rig) and the resolution of the wiimote’s accelerometers. The wiimote managed to get about 50 steps (from pointed up to pointed down) which drove the rig in a clunky sort of fashion. The same range (from up to down) on the rig is handled by 1600 microsteps on each stepper motor. My code in the arduino also needs works to accept instructions at a higher frequency. At the moment it has been optimized to receive a destination instruction by hand for both pitch and roll, and then calculated the acceleration, deceleration and final position relative to it’s current location. All of that work is recalculated each time the python script updates the accelerometer data. Not very smooth.

Today I realized that my shed was getting pretty full of “projects” which pique my interest and then wane. It became clean up day. This was all spurred by the repeated invasion of ants and snails. The ants make a home inside something I am storing (I might find a use for it one day, you never know!), and the snails keep eating my paperwork and notes. Today I had enough and decided to seal all the possible ways in for the creepy crawlies. “It won’t take long” I thought. Now its 9:30pm and all my stuff just made it back inside, all tidied up.

Also I just got delivery of my business cards on Friday. It might not seem such a big deal, but it’s pretty momentous for me. This means that it is all really happening, albeit one step at a time. Insert warm and fuzzy feelings here.

 


3 Responses to “More kinematics and updates”

  1. I want to know more about that

  2. wow why did something like that happen?

    • The jerky motion? I’m pretty sure that the accelerometer microchip inside the Wiimote is very cheap and doesn’t have fine measurements. I think Nintendo would smooth this out via their software to make it seem much cleaner, but I was accessing the raw data stream. There are much better microchips and modules available online to sample this kind of data, and I intend to buy one of those if I ever need to in the future.

Leave a Reply