The 1st (Ferrara) "independent" Tests

Eric Walker, whose views here I find always carefully considered (though I do not always agree with him), has pointed out that of all Rossi's tests Ferrara is the most convincing in the sense that there is not an obvious error mechanism, and the level of documentation is quite good, and the testers are mostly independent (although proved now to be very naive and inexperienced at doing this type of testing). I agree.

The Ferrara tests are in some ways better methodologically than Lugano. So it bugs me that I can't work out a plausible definite error mechanism, and resort to hand-waving "this has not been checked".

There are two tests and a decent explanation of these results must account for both.

REPORT HERE

The table above shows a rough comparison of the two experiments. The most remarkable difference is the different COP value. The authors suggest this is because of the higher temperature in the first test, or because the first reactor had more powder etc.

Higher temperature is unlikely because from the previous Penon tests we know that this device has a COP that remains stable (or decreases slightly) as temperature increases. The Lugano test confirms this E-cat "COP stays roughly constant" observation. It did not show a COP different from 1, but did show with some accuracy that the COP remained stable to within 0.5% between temperatures of 700C and 780C.

Since I do not believe the powder responsible, I need an error mechanism that will give these different results. I could argue that with only two of the 7 testers involved in the experiment, major errors collecting data are more likely in the first test. While that is true I'd prefer a more definite explanation.

The same emissivity issues that afflict Penon are present here. In theory the paint used could have a spectral emissivity that provided and arbitrary positive or negative error in the first experiment. In the second (with possibly different paint) the error could only be positive (or at worst 20% negative) since the temperature is accurately determined at temperature independent of the thermography. In fact that is why the second experiment has much better methodology than the Lugano test. I can't rule out emissivity issues here but don't believe they can account for such a large COP - 2 would be pushing it.

The obvious loophole in this test is the input power. It is better measured then by Penon using a PCE-830. However the details of how the PCE-830 is used are not clear. Nor is the competence of the testers with electrical measurement certain - in the Lugano test they had a few known methodological issues and calculation errors in the electrical analysis that an experienced person just would not have made.

I'm now going to focus on the difference in electrical measurement and show how this could, for a given error mechanism, exactly explain both results.

The second test has X3 COP which is exactly what you would expect if the measurement used for input power at the mains is 1/3 of the correct value. That could happen either through clamp reversal, or through taking the power for a single phase instead of 3 phases. X3 mistakes are awfully easy for inexperienced people using 3 phase electricity.

The first test has X6 COP and measures input power at the device. The difference here is that by having different resistors for the three phases a reversed clamp, or a single phase reading, could easily read 1/6 of the real power instead of 1/3. There is much more room for unusual measurements when the output side of the control box is measured.

So that is my conclusion. It is possible that the additional data from the video of the PCE-830 screen, taken for the second Ferrara test, would shed more light on this.

There are other errors - for example if multiple wires are used inside a single to drive the resistors then the wrong power can be faked with correct measurement because currents cancel. I do not rule that out but it would most easily apply to the first test only.