KryoFlux - Research

2009-09-20

We need to verify whether the board used is capable of continuous USB communication for an extended period of time, that is the first bottleneck.

In order to pursue this goal we had to:

• Learn ARM programming
• Learn this ARM prototyping board
• Learn USB communication
• Learn how ATMEL software library works
• Avoid using any code from other sources than ATMEL and ARM as they usually have mistakes that are only obvious when you have to write code that puts the board on the edge of its processing power

These things had to be studied properly to avoid problems and mistakes. Many of these things wouldn’t have been interesting if we then decided to drop the current hardware, and go with a different type that is more convenient for the cause. The only reason to spend time with them is to produce something. The idea is good.

There is also a possibility that the board’s processing power and its hardware is not suitable entirely for the project - the degree of this is under investigation. If it can be made work at all, we’ll release a version that works with the original hardware, if not we will go with a different type of hardware that is better suited for the task, although more expensive.

Reliability is a serious issue for us, far more than a simple hobby project. These boards are somewhat notorious for being unreliable when used in demanding applications. This is why we needed to completely start from scratch, and make the real deal based on what has been learnt.