Thanks for the info, Sean! I guess I'll just keep track of roll rates with some ADXL202's, and maybe try some cheapo gyros. I don't need really accurate tilt info, just generally where the UAV's at... level, 30 degree bank, 60 degree bank. This UAV needs to be fairly agile, and needs to be able to roll fairly quickly. With some idea of bank, I can predict just when the UAV needs to roll out to be on a certain heading, and how much aileron to use to achieve level flight or a certain bank angle. Thx again! - Robert ----- Original Message ----- From: "Sean Breheny" To: Sent: Thursday, July 24, 2003 9:54 PM Subject: Re: [PICLIST] [OT]: Dynamic tilt angle sensor > Hi all, > > Unfortunately, it is fundamentally impossible to determine your orientation > relative to gravity in a general dynamic environment. Even the aircraft > gyros mentioned earlier will eventually drift away from the correct > orientation if you fly in a continuously accelerated trajectory for a > long period of time. > > To see how this is so, you only need realize that any inertial > acceleration sensor (MEMS accelerometer, bubble level, etc) will show > zero acceleration in free fall. Free fall would be an example of a > dynamic situation where it is impossible to determine the direction of > gravity from purely inertial information. > > The reason why all of these sensors read zero in free fall is that they > in fact do not measure gravity. They measure all the forces on the > vehicle except for gravity. Gravity is special because it is mass > attration and the proof mass in the accelerometer (or the water in the > level) is attracted to the earth with exactly the same acceleration as > the rest of the structure. > > If you can assume that the device is not accelerating and it is in the > earth's gravitational field, then you can use the device to determine > the direction of gravity because you know that the applied forces are in > the exact opposite direction from gravity and the accelerometer reads > those forces. > > I think this problem is usually approached by having rate gyros which > provide orientation change information along with accelerometers that can > tell you your orientation when you are not accelerating. If you then > assume that you are not accelerating in the long term average, then you > use the long term average measured acceleration (which will indicate the > direction of the gravity vector) to compensate for the gyro drift. > > I can tell you right now that this is not an easy task even with > expensive equipment. We've been working with a $17k inertial measurement > unit for our UAV (four rotor helicopter) and we still have difficulty > getting everything right, although it does manage to keep track of its > orientation very well using the above mentioned method. > > A better solution in your case would probably be either a 3 axis > magnetometer (I've never used one so I'm not sure if it can tell you > angular orientation about all three axes) or a horizon sensor as has been > mentioned. My understanding is that horizon sensors simply look for some > line where the IR (or maybe even visible light) radiation changes > abruptly and consider this as the horizon. I could be wrong on this one, > though, too. > > > Sean > > > On Thu, 24 Jul 2003, Tony Goetz wrote: > > > How about using some kind of gravity based sensor (like the bubble, for example) along with a 2 axis accelerometer? The accelerometer would measure vertical and for/aft accelerations. Then it's a simple matter of taking the the components of the vector into consideration. Figure out what angle the bubble/pendulum is at, then subtract the horizontal and vertical forces acting on it due to vehicle acceleration. A fair amount of math to do in a small uC, and you'd need to calibrate it well enough to suit the application, but perhaps it would work. > > > > I don't know, just $.02 from some kid still learning this stuff... > > > > -Tony > > > > > > > > > > > > > The bubble (actually the fluid around the bubble) will be moved around > > > by any acceleration including gravity. > > > You basically can't sense gravity directly without picking up > > > acceleration too. > > > > > > Robert Ussery wrote: > > > > Hmmm... I just had another idea... > > > > How about a bubble level with some IR interrupt sensors > > > along it? The bubble > > > > shouldn't be too affected by acceleration, right? > > > > > > > > - Robert > > > > -- > > http://www.piclist.com#nomail Going offline? Don't AutoReply us! > > email listserv@mitvma.mit.edu with SET PICList DIGEST in the body > > > > -- > http://www.piclist.com#nomail Going offline? Don't AutoReply us! > email listserv@mitvma.mit.edu with SET PICList DIGEST in the body > -- http://www.piclist.com hint: To leave the PICList mailto:piclist-unsubscribe-request@mitvma.mit.edu