I am in total agreement with the statements from Bob. In every simulation that I have conducted using LTspice the system input power is accurately determined by the product of the constant current source DC value and the average DC voltage measured at the node of entry. During my testing I used several different models. In some systems I allowed the resistance from the node to ground to vary according to a sine wave model, while in others I toyed with square wave forms of variation.
I also experimented with additional resistive loads connected effectively in parallel with the DC entry node. Both AC and DC connections were tested for the external node. For some testing I simulated a capacitor that was capable of virtually shorting out the input voltage variations by absorbing most of the AC current being generated by the changing resistance of the modeled cell load. One interesting observation that I carefully observed to be true was that the varying resistance within the cell due to a process such as bubbles forming and breaking actually generates AC power that can be coupled away from the cell under certain conditions. This power can be terminated into an external load and siphons away some of the input power that is supplied by the DC current source. Under this condition the actual input heating power applied to the cell can be less than calculated by an amount equal to that which is lost into the coupled load. This lost power makes the real COP greater than what is calculated. Fortunately, the error is small and only present when an external load is coupled to the cell. There is no indication that any significant load capable of absorbing the cell generated AC power is present during Dr. McKubre's testing. I consider the internal conversion of input DC power into AC power that can be transferred away from a cell such as this to be essentially the same process as seen during the operation of an RF power amplifier. In that case, the device heats up to a temperature that is determined by the difference between the DC input power and the RF output power that leaves the system. The true amplifier heating power will always be slightly lower than what you would expect without any RF conversion taking place. The behavior of a class 'A' RF stage serves as an excellent example of what I am observing in the simulations. Dave -----Original Message----- From: Bob Higgins <rj.bob.higg...@gmail.com> To: vortex-l <vortex-l@eskimo.com> Sent: Sat, Nov 1, 2014 2:42 pm Subject: Re: [Vo]:questions on McKubre cells and AC component BTW, David Roberson and I have corresponded with Barry Kort about the claim that McKubre's measurements were as much as 3% in error due to presumption of constant current and average voltage between samples for calculation of average power. The claimed mis-measurement is attributed to the changing voltage due to the bubbles in the cell rapidly changing the cell resistance and hence cell voltage. Complicit in the argument is the inability of the power supply in constant current mode to adequately slew to keep up with the changes in resistance. Barry claims that reflections setup in the the connecting wires as transmission lines causes dissipation of the time varying component. David and I both did simulations of this setup using SPICE analysis in transient simulation mode, which analyzes the circuit from first principles. In my simulation I used a model for a voltage source in a feedback configuration with a sense resistor to comprise a current source similar to how real power supply current sources are made. Finite slew rate of the voltage was introduced. A lossy transmission line was used between the source and a load resistor, that was modeled as having a sinusoidally varying resistance (+ a constant). The simulated results were compared to that of an ideal current source driving the same load. The instantaneous power waveform was computed in the simulation and its average was taken to get average power delivered by the source to the load. The simulation results confirmed that the use of the constant current value times the average voltage between samples accurately computes the average delivered power. The differences between the feedback power supply model and the ideal constant current source (the presumption) was on the order of ppm, possibly due to the slew effects of the source or just imperfect value for the constant current the power supply sets (due to offset). This ppm difference was far below other errors in any real measurement by McKubre. The "3%" figure for the error in the McKubre's measurements being attributed to use of constant current and average voltage to compute average power in the face of variations in cell resistance appears to be completely unfounded. Barry calculated his solution mathematically including the delta functions that arise from step changes in resistance. He did not go on to simulate his circuit as a check of his math; and I suspect there is an error in his math or in how he has setup his model. Bob Higgins
