Micro Data Acquisition of 3 axis accelerometer

[DNez2001]

First, I am a chemical engineer trying to do electrical engineering work, so I am dangerous . . .

I am trying to develop the smallest possible data acquisition system for monitoring and logging data from a 3 axis accelerometer taped to a sports object (I would rather leave this undefined at the moment). I am trying to minimize the profile and the weight due to aerodynamic considerations. The assembly will need to be glued to the sports object so that it doesn't fall off during the high acceleration it sees, and it needs to be self powered to support the logging of the data on the EEPROM section of the microprocessor. It needs to be able to store data for all 3 axis' for about 10 sec at about 100-200 points per sec. I am considering:

A PIC12LF1822 microprocessor (4.9 x 6 x 1.75mm, wt unknown, but probably less than 2 g)
A BMA220 digital 3 axis accelerometer. (2 x 2 x 1mm, wt unknown, but probably less than 1g)

The battery and the circuit board to wire these together, as well as a 4 wire connection for USB interface for data acquisition and programming are going to weigh much more than the components. Looking at the specs, it appears that this will draw about 2 mA running at 16 MHz.

Any advice and counsel on how to approach this project would be appreciated. I am not planning to conquer the world with this, I am only trying to build a few of these for enhancing sports performance.

[Soeren]

Hi,

[...] The assembly will need to be glued to the sports object so that it doesn't fall off during the high acceleration it sees,
[...]
Any advice and counsel on how to approach this project would be appreciated. I am not planning to conquer the world with this, I am only trying to build a few of these for enhancing sports performance.

If you're truly "only trying to build a few of these", I don't see a reason for the "I invented the golden egg and nobody's gonna steal my idea" attitude.
It's really hard to help people who's that secretive. Imagine I asked you to help compose the chems of something without giving you half a clue - I think you'll agree that engineering is a far cry from ESP 

My first concern with what you wanna do, which reminds me of the data logging GOLF ball, is that you wanna glue it on the outside of the "projectile" - it will give the electronics a hard time and it will change the characteristics of said "projectile", so what you pick up will be influenced by the physical bits of the electronics.
Another concern is that I sense, that the carrier (ball or whatever) might be given a good kick or a whack with some sort of "stick", and that's a real challenge, making it survive that.

My best advice is to start with spilling your guts. Then perhaps a solution could be found.

Good Luck with the project whichever way you decide to take it.