Cool Engineering

Info on some cool engineering projects

Tuesday, May 13, 2008

Clocking a POV demonstration

With open day coming up again at Uni (a couple of months away still) I've been brainstorming different ideas which would have a good visual aspect to them. One such is a fairly simple line following robot - the board for which has been made and I'm slowly assembling. It will be based around an AT Tiny2313. Realistically is could have been implemented using an AT Tiny13 - but I wouldn't have gotten huge amount of feedback as to what state the PID was in so it would have been more difficult to design the control algorithms. Now we have extra LEDs and pins to spare.
The other project I've been pondering is a POV (Persistence of Vision) toy. Now according to Wikipedia is seems the term Persistence Of Vision is rather inaccurate as it was based on a flawed concept of how humans interpret motion. It seems that thinking about persistence of vision treats the eye as more of a crude camera - wherein it is more complex all together
. Either way basically we want to quickly move a changing line of LED's, thereby allowing us to write messages or display patterns. Some people have put together this device as a handheld writing device which would be pretty cool for open day - one concern is that someone will eventually get wacked in the head - or we have some sort of issue like with the flying Wii remotes. Another common implementation is to have a rotating line of LEDs and to change the text based on the position of the rotation arm. Additionally it'd be kind of cool to have it wirelessly updatable, and we would need some sort of sensory feedback as to what point the arm is at so we can get all our timing right. For doing that I'm currently thinking a photodiode and either an IR LED or a black/white printed piece of paper will work similar to the line following robot - another common technique seems to be using a hall effects sensor and a magnet. In code this will be interrupt driven and continually dump out the value in a free running hardware timer before resetting the value of the timer.
For planning purposes it is important to get something that actually swings - and does so at a descent enough speed. Trialling a few different small DC motors I clamped the motor and a vice attached a small metal bar (with some batteries connected to simulate the final product). To calculate the speed of rotation I affixed a bright torch (Surefire - yes it was overkill!) and on the other side positioned a photodiode which was connected up to my CRO. Now it was a simple matter of reading the waveform off the CRO and to my delight we were doing 33 revolutions per second - which considering PAL is 25 frames per second - should be enough. I was only driving the motor at 3V and it was rated to 6V (so we have some room to move!)

Sunday, May 11, 2008

Study trip to Brisbane

Last week marked I conducted a whirlwind 2 day study trip to Brisbane involving visiting CSIRO, QUT and UofQ. Several points that stood out from all of this:
- CSIRO is really cool and produce amazing autonomous stuff from submarines to bobcats!
- QUT has a really cool UAV department and some very switched on PhD students
- UofQ has some great mechatronics facilities
- All of the people I talked to were without exception really friendly, made time in their busy schedules and offered really helpful PhD advice
- It seems impossible to get free WIFI in Brisbane.