Re: NXT Synchro Drive
"Chris Phillips" <[email protected]> Thu, 7 Dec 2006 07:03:22 GMT
| Newsgroups | gmane.comp.hardware.lego.robotics |
|---|---|
| Organization | None |
| Message-ID | <[email protected]> |
In lugnet.robotics, Rich Thompson wrote: > I have been playing with the turntable from the NXT kit to try and > build a synchro drive train. It got a bit complicated but it works. I > posted a video at my blog: > > <http://www.robotthoughts.com/index.php/lego/archives/2006/12/06/lego-mindstorms-nxt-synchro-drive/> > > or just: > > <http://www.robotthoughts.com/> > > Have Fun! This looks pretty cool, and I can well respect any attempt to design a cool device out of LEGO, but I wonder how necessary something like this is with the NXT? It seems to me that the new servo motors with their integrated rotation sensors are incredibly accurate for navigation, especially since the firmware is capable of precisely matching the speed of the motors. The design of the servo motors (as depicted in the NXT booklet) has the rotation sensor geared almost directly off the motor shaft at the head of the internal reduction geartrain, so it has extremely high accuracy. The documentation claims it has a resolution of one degree of output shaft rotation! The demo robot, built from the Open Me First set of parts and run with the Demo program that comes pre-loaded on the NXT, is an amazing demonstration of this capability. This robot moves straight forward a few feet, then turns 180 degrees, returns to its original location, and then turns 180 degrees again. On a level surface, I have found that it is capable of returning to its original location within a few millimeters! Navigation by odometry was something I had long tried to accomplish using <http://www.drvegetable.com/pic_franken02.html the RCX and a pair of Rotation Sensors> without ever coming close. With odometry so greatly improved on the NXT, is a synchro drive platform really necessary? Just wondering...