Any and all help/suggestions appreciated. Need to detect leg (human leg) motion during gait and based on tilt/inertia/position would like to signal a single output (motor). Would I need a gyro/accelerometer combo. or just a gyro or just an accelerometer. At this point I just need ideas/suggestions, pros/cons. Also can this be done with an Arduino Uno using the motorshield. Any input appreciated, at this point just want to be pointed in the right direction. Thanks for your help! [/color][/color]
Can you tell us more about the purpose of the project and possible constraints?
You probably need to give a bit more information. Which bit of the leg are you interested in the femur, the tibia or both?
My first thought is that you would probably want sensors on the hip plus the main bones down to the lowest one you are interested. Since the bones are jointed maybe you only "need" accelerometers, but perhaps it would be easier to have a combo - others will know a lot more.
I would like the sensor(s) to be placed on the lower leg (tibia), above the ankle and below the knee.
You are still being a bit sparse with information on what you want to achieve.
If you are interested in the movement of the tibia then don't you also need to have data from the hip (to know the persons forward velocity) and the fibia (since it will be throwing the lower leg forward)? Or is there some reason why you can look at only the acceleration of the tibia in isolation?
Sorry, I'm not sure at this point exactly what the function of the motor will be. I would like to have the unit "self contained" on the tibia only. I'm only concerned with the motion of the tibia, not the foot, or above the knee, during normal walking. 
Hope this helps.
It still isn't clear what you are trying to detect. The tibia is not moving in isolation (at least not normally).
Try again to describe your project, in non-computer related terms. "I'm trying to detect the speed/acceleration/amount of rotation/etc. of the lower leg as a person walks/runs/kicks a soccer ball/sits and drinks beer. I want to buzz them when they are too slow/activate a fan when they go too fast..."
Christina Stohrmann of the Swiss Federal Institute of Technology (ETH) in Zurich has attached sensors to a runner's hip and foot to distinguish between the contrasting running styles of novices and those more experienced. The system was presented at the Ubicomp conference in Beijing China.
The "sensors" are being described as Inertial Measurement Units (IMUs). Each IMU contains an accelerometer, gyroscope, and magnetometer to measure strength and direction of magnetic field, similar to those found in an iPhone.
You might want to find out what this about. It gives an indication of what other people are using. It is interesting also that that there is a sensor on the hip and foot but none on tibia and fibia.
It is difficult to understand what you are trying to get by looking at the tibia in complete isolation even from the hip. As PaulS said you should try and explain in plain English what you are trying to do.
Ok, let me simplify it. I do appreciate the responses. Just imagine a velcro type cuff w/ a box just below the knee. When someone walks (no stairs, no running, no jumping) I would like to have sensor(s) in that box that tell me how that box is moving. Obviously, with all due respect, I know the lower leg is connected to the rest of the body.
I'm only concerned about the motion of that box in the forward linear direction. That motion will also be up/down, side/side and starting/stopping in a forward direction as the person walks. I'm not sure yet what any outputs would be for now.
I've looked at other posts about accelerometers, like the mesmic 2125 in particular, which seem like it would allow me to measure this forward linear motion. Some of the posts mention that if you are trying to measure/monitor motion in one direction and the sensor is rotated that can "throw off" the readings. Is this true? Would it have to be used with a gyro or something else? Or would just a gyro work?
I'm interested in trying figure this "box" out.
As of now, I'm envisioning using an Arduino uno w/ some kind of shield (protoshield) and accelerometers/gyros/or whatever else.
Again I'm a newb and haven't purchased any components yet. I am fascinated by the whole Arduino concept and want to see if I can figure this out.
I greatly appreciate any and all replies.
Dave_RI:
Ok, let me simplify it. I do appreciate the responses. Just imagine a velcro type cuff w/ a box just below the knee. When someone walks (no stairs, no running, no jumping) I would like to have sensor(s) in that box that tell me how that box is moving.
This is starting to get a bit more tangible.
How accurately are you trying to measure it, and over what sort of time/distance, and under what conditions? I mean, are you just trying to gauge how long somebody's stride it, or tell them whether they've finished their marathon yet, or produce an inertial navigation system, or what?
Although you seem to be mainly interested in movement along one axis, it will probably be very hard to do this with any accuracy unless you can control (or determine) the orientation of the sensor. We have a similar problem with motorsport datalogging because it turns out that cars going round a circuit are rarely pointing exactly where they are going. The common solution in that case is to use high resolution GPS to compensate for cumulative errors and tilt/yaw/roll effects. Depending what you're trying to achieve, you may need to do something similar. Hard to tell though, because it's still pretty unclear just what you are trying to achieve.