Also,

As I try this more, I realize that as the frequency approaches zero, I start to get a non linear counting result from the fix 5KHz counted stream. That is something else that I have to re-factor in...

hmm.. This is more problematic than I thought. There has got to be a simpler way to do this without reinventing the wheel. I must be stuck on stupid......:(

To recap, I have a timer free running at 5Khz and incrementing a counter after each overflow in an array every time the counter reloads, when the input pulse hits, it resets the count overflows to zero. I'm using that to see if I can correlate input timing pulse to a fix reference that can be used to create a function that can output the correct speed data.

Does that sound like what you expected Per? In the time domain?

Or am I still on the wrong track?