Desp,
If the guestimates the TDC controller produces were any good, we'd need not have this discussion. There are simply to many variables that the controller method depends on that are nor really know for all time points.
My believe is that the accuracy of your data is not much better than what we get from the TDC, so the trust in the results wouldn't improve, sadly. Just look at the noise (jaggedness) of the calculated power versus time plot. There is probably a way of estimating how many independent readings you need to arrive at a reliable average. And that still leaves you with the question of What average you've got?
So, I think Eric's latest baby has been born just in time for you to get onto the data-logger train. May be you could even find / make another hole for an additional sensor in your store? I'd certainly be there for the launch party
Klaus
