If it does not show up, how could it be measured?
Actually, this can be accomplished in an interesting sort of way with the program. On occasions I intentionally restrict the range of data used in the optimizer. If I want to concentrate upon the rising edge fit, then I do not include the later data in the LMS routine. The output of my optimized variables contains one that corresponds to the input power. This usually matches up to within .2 watts or so and I know at that point that the fast acting effects are taken into account. Now, as time progresses and you look at the error data, you will see any tendency for the excess power to change. It could ramp up or down or stay the same. If instead you are interested in the best overall long term fit to the data, then you can restrict the optimizer input to include the later data to the end of the run if you wish. For this type of test, I leave all of the time variables as well as the initial power set (Kint) fixed and just optimize the Pin. This would muck up the match for fast acting processes, but concentrate on the long term behavior of the cell. Once optimized, the Pin will indicate the effective input power which includes any excess power being generated by LENR. If the error is now flat and you see that Pin is not what you actually know is being applied to the cell, then you have something going on. This simulates what you are thinking I believe. It allows me to concentrate on short term effects or long term effects depending upon my expectations. Dave -----Original Message----- From: Daniel Rocha <danieldi...@gmail.com> To: John Milstone <vortex-l@eskimo.com> Sent: Wed, Feb 6, 2013 3:36 pm Subject: Re: [Vo]: MFMP Null Result Can't you simulate a few types of dummy systems with extra heat where the extra heat would not show? 2013/2/6 David Roberson <dlrober...@aol.com> That is what should be showing up as time progresses. If the calibration values are determined by the faster acting phenomena, then a set of values is obtained that is accurate for fast moving changes. The time domain fit to power steps demonstrates that this is happening and fitting the calculation very closely. Now, if the slow later things come around, then the long term watching of the calculated power would show an increase if excess power is generated or a decrease if some form of endothermic action is happening. My program fits fast changes on the rising edge and then becomes flat at a value that depends upon the quasi static calibration points. This type of procedure should be powerful in demonstrating LENR activity. Good questions Dan. Dave -- Daniel Rocha - RJ danieldi...@gmail.com
