by hivemind » Mon Aug 14, 2006 2:36 pm
by hivemind
Mon Aug 14, 2006 2:36 pm
CaptKill4Fun wrote: Lots of talk, but no demos, no posted code, nothing but talk ...
Hmm, is it just me or do other find this post ironic?
And besides routine demos that I test with no controller/my computer, I also use bluetooth with a cell phone which enables me to use two gyros and a three axis accelerometer. I have two hole cutouts in my chest so i can adjust the gain of the gyros at any time. It is also possible to change a speed varaible via the phone, and have a nice gui so that i can have a huge variety of moves so unlike a ZigBee linked PS2 controller and R/C i am not limited to a few moves or many combinations.
As for the monitering, what? The gyro settings and accelerometer readings would be of utmost importance, lets say for example your playing robosoccer and you keep getting knocked down by other robots. (because:
CaptCrunch4Fun wrote: I don't see a need for a side-2-side gyro
) But why not have a setting that you can switch on via the phone that could up the gain of the gyros if the x (or y if you had that port...) axis of the accelerometer is hit too hard (ie: above 1g). This could give your robot ample time to react next time he was hit, and then, even if he did fall over you still have the port open, so using a three axis accelerometer it can stand up on his own. (wow)
Even Pev runs bluetooth via a PDA, and i would guess that he would have the robot sending back a constant status report including the accelrometer readings, gyro gain,perhaps eye colour, and why not? It is a highspeed connection for the robot to a cpu which can handle much more complicated algorithms. You could be using the accelerometer and gyros anti-derivative and trace the robots position from a starting point. Could you do that with R/C? I shudder to think of the lines of code and the lag time of the robot if you tried.
So CaptKill4Fun, where are your
demos? Oh wait, they are amazing, in fact, great job on the shovels second try. The finger.
No really.
CaptKill4Fun wrote: Lots of talk, but no demos, no posted code, nothing but talk ...
Hmm, is it just me or do other find this post ironic?
And besides routine demos that I test with no controller/my computer, I also use bluetooth with a cell phone which enables me to use two gyros and a three axis accelerometer. I have two hole cutouts in my chest so i can adjust the gain of the gyros at any time. It is also possible to change a speed varaible via the phone, and have a nice gui so that i can have a huge variety of moves so unlike a ZigBee linked PS2 controller and R/C i am not limited to a few moves or many combinations.
As for the monitering, what? The gyro settings and accelerometer readings would be of utmost importance, lets say for example your playing robosoccer and you keep getting knocked down by other robots. (because:
CaptCrunch4Fun wrote: I don't see a need for a side-2-side gyro
) But why not have a setting that you can switch on via the phone that could up the gain of the gyros if the x (or y if you had that port...) axis of the accelerometer is hit too hard (ie: above 1g). This could give your robot ample time to react next time he was hit, and then, even if he did fall over you still have the port open, so using a three axis accelerometer it can stand up on his own. (wow)
Even Pev runs bluetooth via a PDA, and i would guess that he would have the robot sending back a constant status report including the accelrometer readings, gyro gain,perhaps eye colour, and why not? It is a highspeed connection for the robot to a cpu which can handle much more complicated algorithms. You could be using the accelerometer and gyros anti-derivative and trace the robots position from a starting point. Could you do that with R/C? I shudder to think of the lines of code and the lag time of the robot if you tried.
So CaptKill4Fun, where are your
demos? Oh wait, they are amazing, in fact, great job on the shovels second try. The finger.
No really.