Vehicle Tilt Control?

Lean

I am building a three wheel electric vehicle which has the ability to tilt with·mechanical input on corners but·I want it to be able to do so automatically. I understand that it will require both gyro and accelerometer sensing, a control circut, a controller, software and input, plus a servo motor. This will all drive a jackshaft forward and reverse. I have done the mechanical stuff but the electronics are·beyond my ability. Any suggestions or help would be greatly appreciated.

Franklin

I would say your first job will be figuring out what physical forces you need to sense to activate the movement and how you will limit and control that movement. This is not the electrical part but it will help in deciding what you need. Let us know what you come up with.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Stephen

Lean

I'm on my handheld so this will be brief. The side forces in going around a corner in the vehicle (tilting trike) is estimated to be 2gs. The servo motor will need to remember the trikes upright or center position. So I'm thinking that as the vehicle go goes into a turn and begins to tilt due to g forces an accelerometer will sense the force and send a small voltage to a controller that will signal a servo motor to run the jackshaft and keep the trike upright. Then as the force is reduced coming out of the turn, the jackshaft returns its attachment nut to the center or home position. Since this is all very dynamic a gyro is needed to help stabilize movements - or so I'm told. I'm not interesed in dealing with starting or stopping acceleration forces except in how they may affect the tilt control - this may be where the gyro comes into play? THANKS

Franklin

OK, as I see it you can use the accelerometer to sense 'down' and as the trike turns compensate to keep the axis straight 'down'

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- Stephen

parsko

Lean,

The only way to do it reliably is to include a gyroscope. An accelerometer will only work on static or very low-G applications. Once you get something moving, such as a trike, the acceleration (which you were using to sense position while static) will give you a false reading (due to the G-forces induced while turning). A gyroscope will give you your tilt, which you would then use to "remove" that much acceleration from the accelerometer reading, giving you your absolute acceleration (aka the amount of G-force your servo's need to compensate for).

I think you could do it with only an accelerometer if you integrated over time, but you'd end up with error stack-up which I think would lead to a diverging control system.

-Parsko

ChrisP

Accel and decel should have little or no effect on the servo system, the only time these would really effect the dynamics at all would be when experiencing these forces during cornering, and if I understand your intention correctly (to add stability to a 3 wheel vehichle) then accel and decel still have little to no impact on what your trying to achieve.

Motorcycles are inherently stable during cornering because cornering G forces are experienced by the suspension almost completely as "down" forces instead of lateral load on the suspension due to the bike being leaned over the in corner. Whatever angle a 2 wheeled vehichle is to the ground during normal striaght line or cornering operation the subjective "down" is nearly always straight through the vertical centerline of the chassis, with very little angular offset.

Sooo all that being said.... From my point of view, the important thing to your app is "subjective" down... Subjective because it has very little to do with what is actually down, other than the fact that a combination of gravity and lateral cornering forces combine to alter the perception or target angle to compensate. An accelerometer with sufficient damping or filtering algorythms may be possible, another approach may be to look into RC helicopter solid state gyros to more accurately determine proper lean angle.

With a gyro, it should be simple.. Just have your servo system react in an opposite fashion to your gyro's swing from calibrated vertical to maintain an upright position. As lateral cornering forces build your gyro will think its leaning over and the servo system will lean your chassis in an attempt to maintain a vertical position.

All in theory, and lots easier said than done but should work just fine.

Lean

Thanks for the responses everyone! I will look into a model helicopter gyro unit. My problem then will be in finding the proper control unit to put between the servo and the helicopeter's gyro output signals.

Javalin

Lean

I wouldn't touch a heli-gyro - except with a big stick....! I bought one thinking that'd be ok - dunno whether its just mine but couldn't get anything usefull out of it.

Look at Sensors->IMU category at www.sparkfun.com

Bought some of their stuff recently - really good.

James

Lean

Thanks for the info James. I see that "Sparkfun" has IMU Combo Board with both gyro and accelerometer. If I can figure out how it interfaces with an encoded servo motor, I may go ahead and try one.