STEORN DEMO LIVE & STREAM in Dublin, December 15th, 10 AM

[teslaalset]

Of course, it would be nice to explore also the offset question but my main idea was to make a graph where we will see the criss crossing of the two best-fitting lines. It may require hundreds of thousands of points, maybe millions. I still don't have Excel 2010 and for now won't be able to open any such file. At present, like I said, I can hardly handle 32000 data points with some rudimentary Excel 97.

Isn't that what the trendlines do for you?
I noticed you used those already in your graphs.

[Omnibus]

Like I said, there is a chance of floating point errors here. The change is just about 0.5*10^-12, this is unimaginably small. There is no way to tell, unless someone derives it from first principles. 

I guess I'm missing something here, too many posts to read, so I read them in hurry, and the question is, why is broli getting a positive slope ? Is he using a different formula ?

About Excel, it can plot only 32k points, even Excel2007 has this limit. But you can always reduce the points by increasing the time step, so you arrive at a full cycle early. And you can remove the markers from the plot to get a nice clean thin line. (right click on the line and select format data series)

Now, I don't think @broli can determine the slope of the energy-time curve in Mathematica. Maybe it can be done but at this point I don't see how, unless he resorts to generating lists of data to proceed the way its done in Excel (to do that in Excel seems easier and more transparent). What he does is integrate the momentary powers over a certain period of time. To determine the slopes we're talking about one needs to have the values of a series of such integrals plotted as a function of time and find the slope of the resulting curve (the best-fitting line through that curve having fine structure). So far this can only be done conveniently in Excel, as far as I can see, obviously only numerically. I don't see at this point what the analytical expression might be for such slope resulting from continuous curve stemming from integrals over time. That'll be great if someone can suggest a way to do it. That indeed would be the ultimate proof that standard theory of electricity contains inherently OU.

[Omnibus]

Isn't that what the trendlines do for you?
I noticed you used those already in your graphs.

Yes, they do, of course, and maybe what am asking for is an overkill but I thought it might be nice to see the criss crossing in front of your eyes and not only rely of the fact that the trendlines will definitely cross somewhere in the distance. But, maybe you're right. What's done so far will do. One thing to address here is the point @Omega_0 is raising about the floating point error. There should be an assessment of this error. Do you have any idea how this can be done?

[Omega_0]

@Omega_0,

Be assured I will do anything in my power to bring these new ideas and findings to the attention of any influential, competent people that I am able to come in contact with. You may realize that this is a difficult task and must be done in a strategic way so that it will not be squashed. The first I'm planning on at this point is to convince reputable labs to reproduce it independently. I will let you know how this develops. In the meantime I'm again strongly encouraging you to submit a manuscript describing your results as a preliminary note to a scientific journal as prestigious as possible.

Also, I would be willing to give a joint presentation with you on all of these findings. I'm planning to be in Europe by the end of the Summer and we may try to arrange such seminars. You may want to suggest universities for that matter and I will contact them and try to arrange colloquia at the appropriate departments to discuss all these findings.

It will be great if even one lab can reproduce it. But I'm sure you can convince them. I will post detailed build instructions soon, and I'm posting the data again in text format to end the Excel2007 dependency.

[teslaalset]

Yes, they do, of course, and maybe what am asking for is an overkill but I thought it might be nice to see the criss crossing in front of your eyes and not only rely of the fact that the trendlines will definitely cross somewhere in the distance. But, maybe you're right. What's done so far will do. One thing to address here is the point @Omega_0 is raising about the floating point error. There should be an assessment of this error. Do you have any idea how this can be done?

As long as 2 point linear trendlines are use on graphs that represent exactly N*360 degrees, these are very reliable and as accurate as Excel itself.

Floating point errors can be calibrated by just using 0 Volt input.
Those trendlines should give you exactly zero angle and zero offset.