Posts by Ascoli65

    In this last day of suggestions to the Team Google, I'd like to stress that a mandatory goal of the next investigation on CF is to avoid any negative outcome. A negative is not useful to anybody. Reporting a failure to measure any excess heat will be criticized by both the mainstream scientists, which could say that there was no need to waste money and time on this largely predictable outcome, and by the LENR fans, who will find many reasons to justify the negatives, like, for example, those reported in the recent ArXiv paper (1): "… possible reasons for negative results of 'cold fusion' experiments published recently by the Google-organized scientific group are given."

    If Team Google is aimed at obtaining some excess heat, they are going to dig the umpteenth hole in the water and their work will be basically ignored by mainstream scientists and LENR fans and, this time, also by Nature.

    It's necessary that their next paper begins with a positive declaration, something like: "We have succeeded in replicating the fundamental phenomenon observed by Fleischmann and Pons since 1986 and which led the research in Cold Fusion for the next decades. This phenomenon, called "positive feedback", has been largely discussed in many of their articles and is well documented in their 1992 video, which clearly show its dramatic effects on the electrolyte. We have been able to accurately replicate this phenomenon at will on a regular basis and we are now able to explain all its characteristics and the real source of the excess heat claimed by the 2 pioneers of Cold Fusion".

    If Team Google wants to attract the attention of the mainstream science and avoid harsh criticisms from both CF supporters and deniers, they should carefully select the goal of their next study in order to be sure that it will be successfully achieved, that is that they will get a positive outcome, replicable by any other laboratory. IMO, the only goal which offers a 100% guarantee that will be achieved is the replication of the "positive feedback" phenomenon documented in the F&P video of the "1992 boil-off experiment".


    Why do you think it took Mizuno years to get it to work, and about 3 years to get back to the high power levels he first measured in 2016, after years of seeing nothing?

    Because this is a likely frequency at which it could happen that a researcher eager to make a sensational discovery makes a providential, albeit inadvertent, major error which allows him to come across these extraordinary results and, since he strongly wants to believe that this erroneous results are real, he avoids checking the experimental set-up and data with the care required in these occasions.

    By the way, the visitor -- like me and everyone else -- brought his own power meter and confirmed the input power. So now we have at least 5 meters agreeing that the input power is 50 W, not 300. But I am sure you will continue to claim all 5 meters were wrong. I suggest you find an authoritative document describing how power measurements of an ordinary resistance heater powered by an ordinary power supply can be measured incorrectly to this extent by 5 different meters, including plug in and clamp on types, including one that cost $16,000, all to the exact same extent. Please, show us how that can happen.

    Why don't you start showing us how it is possible that, in the spreadsheet you published in 2017 (1), the "input power" values measured directly by the $16,000 wattmeter were replaced by the product V*I (2,3), please? It would appear like a deliberate manipulation of data. Could you give us an alternative explanation?

    You know, publishing manipulated data would not be the best way to improve confidence in the reality of the data supporting extraordinary results like those claimed for the R19 and R20 devices.

    (1) Mizuno : Publication of kW/COP2 excess heat results

    (2) Mizuno reports increased excess heat

    (3) Mizuno reports increased excess heat

    While you're at it, don't forget to look at:

    K. L. Shanahan

    "Comments on ‘Thermal Behavior of Polarized Pd/D Electrodes Prepared by Co-Deposition’",

    Thermochim. Acta, Vol. 428, pp. 207-212 (2005).

    Hi Kirk,

    thanks for the tip. Your paper confirms that the 2004 article from Szpak, Mosier-Boss, Miles, and Fleischmann is an important milestone in the development of the CF affair.

    I think that the "positive feedback" played a central role in this story, therefore I agree in particular with this phrase of yours: "As well, both the ‘proof’ of a positive feedback relationship and that of a heat-after-death condition are singular examples. Replication is required to prove the observed events are not random coincidences. Replication would also demonstrate control of the effect, which is missing to date."

    The Google initiative is a unique opportunity, probably the last one, to replicate the F&P experiments, in particular the "1992 boil-off experiment", and deeply investigate the mechanism of the "positive feedback". Undoubtedly, their experts are able to replicate this phenomenon on a regular basis and they will be able to solve the CF cold case once and for all.

    This is strictly my opinion, but I do think you and THH are making a good argument in defense. Those older Pd experiments are extremely well documented in the literature, and there are many of the old guard left who I think would be more than happy to offer TG their assistance.

    Furthermore, some of them still own the original F&P's open cells.

    Here a couple of interesting shots btw. Jean-Paul Biberian brought along one of Martin Fleischman's original vacuum insulated calorimeters. Here's some photos.

    Here,s the inner electrode assembly

    And here (lying on the table) is the twin-wall outer container with its half-silvered upper section.


    To a cold fusion experimenter these are 'holy relics and the chance to handle them pretty powerful medicine.

    Hopefully, they could lend these cells to the Team Google to be tested again, or to be used as model to make similar devices.

    Anyway, as you said, the old F&P devices and methods are so well documented in the available literature that Google experts will have no problems in making a Dewar-type calorimeter capable to reproduce the "positive feedback" phenomenon, which is at the basis of the most important F&P's claims. The golden rules are also listed in the paragraph 2.5. "Dewar-type calorimetry" of the SPAWAR article published in 2004 (1).


    You might as well expect THH to admit that when 4 different power meters show that a resistance heater has 50 W going into it, some of them clip-on, some direct, and they all agree to within less than 1%, that proves there is 50 W going into it. Not 300 W. THH and the other pathological skeptics will never, under any circumstances, admit this is true, or that any other cold fusion claim is true. They will continue to insist there might be an error measuring input power, and what looks like 50 W might actually 300 W. Of course they cannot actually cite any error, for this or any of their other crackpot claims, but that is no impediment. The laws of physics, common sense, reality and what actually happens means nothing to them, or the editors at Nature. Politics, power, money and primate emotions are all the only things they know. These are not rational people. They are not doing science. Their goal is not to discover new knowledge, or improve the prospects for humanity, or prevent global warming, or even to satisfy curiosity. Their only goal is primate politics: dominating others as Alpha Males. Being "right" by intimidation, posturing, hooting and throwing sticks and feces. Primate behavior is as much a part of human behavior as it is in chimpanzees or gorillas.

    Why do you expect people should believe the Mizuno's extraordinary results without even looking at the original spreadsheets? The only spreadsheets available on internet, those referring to the 120 W active and control tests carried out in May 2016, contain many incongruities. The most serious being the cancellation, in the active test spreadsheet, of the powers directly measured by the wattmeter and their substitution with the results of product V*I (*). Results claimed for R19 and R20 are even more astonishing, and, to be believed, Mizuno should first reveal who modified the 2016 active spreadsheet, while providing a reason for doing so. A good reason, something capable of removing any little suspicion that the values of the original direct measurement of input power acquired by the wattmeter were deleted because they showed a power level double compared to that obtained by multiplying V and I.

    (*) Mizuno reports increased excess heat

    Correct me if wrong, but right now my take-away after reading hundreds of comments, is that the LENR communities message to TG is that:

    "There are no experiments we can recommend with a guaranteed 100, 80, or even 50% chance of seeing an LENR effect. ..."

    I'm sorry, Shane, but IMO this is absolutely wrong.

    The "positive feedback" is a phenomenon which F&P have claimed to have seen in their experiments since 1986 (1) and many LENR researchers at various labs have claimed to have reproduced it at least until 2002 (2). So it is 100% guaranteed that Team Google has the resources, time, and talent to successfully reproduce this phenomenon and all its effects.

    Consider, please, what was just said by:

    Hundreds of researchers at major labs such as Los Alamos, China Lake and BARC published peer-reviewed in major journals papers making this claim. These experiments were every bit as convincing as Mizuno's is, from a scientific point of view. The papers are definitive and irrefutable. No skeptic has ever discovered any significant error in any of them. Many of the researchers were far more important than Mizuno, with more credibility. For example, the head of BARC and later Chairman of the Indian Atomic Energy commission published papers. ...

    All of these researchers replicated the PdD electrolytic experiments of F&P by using bulk cathodes or the co-deposition method. Considering only these last experiments, the number of confirmations and endorsements is really impressive, as recently reported in the Synopsis of Nuclear Reactions in Condensed Matter (3):

    - 60 peer-reviewed publications (the last in 2018);

    - 38 U.S. plus 22 non-U.S. authors from 12 countries;

    - 25 U.S. plus 13 non-U.S. institutions.

    Which other CF experiment can boast such a general consensus among the LENR scientific community?

    Actually, it's true that only a very low level of excess heat was claimed by using the co-deposition method, but the cell temperature was not increased up to the boiling point, as was said by F&P it was necessary to obtain power densities comparable to those typical of Fast Breeder Reactors. The Team Google can easily overcome this limitation by extending the temperature range up to the boiling point and letting the electrolyte to completely evaporate as happened in the "1992 boil-off experiment".

    Once Google's experts have procured and properly instrumented a set of open electrolytic cells, as similar as possible to those used by F&P in 1992, they can test a wide spectrum of different specimens, electrolytes and operating conditions, so that they will be able to reproduce the "positive feedback" phenomenon which led the multi-decadal research activity of F&P and of their emulators and epigones.

    100% guaranteed.

    (1) Team Google wants your opinion: "What is the highest priority experiment the LENR community wants to see conducted?"



    Using a Dewar-type electrochemical cell/calorimeter, it
    was shown that the rates of excess enthalpy generation using electrodes prepared by the Pd/D codeposition technique were higher than that obtained when Pd bulk electrodes were used.6

    I'd like to point out the above sentence which shows that in 2008 F&P open cell calorimetry was still considered very reliable and effective, even by the experts at SPAWAR.

    I also suggest to look at reference:

    6. S. Szpak, P.A. Mosier-Boss, M.H. Miles, and M. Fleischmann, ‘Thermal Behavior of Polarized Pd/D Electrodes Prepared by Co-Deposition’, Thermochim. Acta, Vol. 410, pp. 101-107 (2004).

    Figure 1 of the aforementioned article (1) shows the Dewar-type calorimeter used in this work carried out in 2002 (date of submission to Thermochim. Acta). This open cell is identical to the 4 cells used in the "1992 boil-off experiment". Therefore, replications of the co-deposition tests performed at SPAWAR could be easily coupled to the reproduction of the boil-off test performed with other more simple cathodes, made of bulk Pd or even Pt. Most of the lab equipment would be the same.

    Furthermore, it would be interesting to understand why two phenomena such as "heat after death" - which was claimed to occur "at high temperatures (in the vicinity of 100°C)" (2) – and "positive feedback" - which was put in relation with "the rapid increases in temperature towards the boiling point" (3) – were also been claimed to be obtained at much lower temperatures, not greater than 320 K (47 °C), as shown in Figure 2 of (1).

    So, why not, replication of both the "1992 boil-off experiment" and the "2002 co-deposition test" would be a good way to optimize time and resources devoted by the Team Google to this initiative.




    What is wrong with this argument?

    Best wishes, THH

    PS - before ascoli jumps in. I'm happy for the initial replicable (and replicated) open cell boil-off electrolysis experiments from FP to be tested as well, if that is deemed the clearest evidence of LENR in these D/Pd experiments. Personally I'd rather explore closed cell work.

    I agree with your arguments and, after your PS, I agree even more.

    An important and well funded initiative such as Google's can afford to replicate both the types of F&P cells, open and closed. Two is better than one.

    However, I stay on my idea that only the "open cell boil-off" test will allow the Google testers to deal with the objections such as the one just raised by Zeus46 : "… some electrodes seem to produce heat, and others don’t. And no-one knows why."

    In fact, when the Team Google will reproduce the same behavior shown in the F&P video (1), and there is no doubt they can easily do it, the objections will be limited to the interpretation of the experimental data. There will be no more room for other objections.

    Consider for example a possible objection on the crucial point of "loading", that you have reminded in your post. The Team Google will be never able to attain long lasting and measurable loading values such as those which, in the F&P interpretation of their results, allowed them to obtain excess heat levels above the noise threshold. The reason is that, by their interpretation, the adequate loading is obtained thanks to a dynamic and short lasting phenomenon, they called "positive feedback" (2). This "positive feedback" is intimately connected to the boil-off phase, therefore if the Team Google wants to avoid any objection on loading inadequacy, they must reproduce the "boil-off" transient and, most importantly, they must demonstrate to have reproduced it. The only way they can do it is by reproducing the F&P's "1992 boil-off experiment".

    (1) Team Google wants your opinion: "What is the highest priority experiment the LENR community wants to see conducted?"

    (2) Team Google wants your opinion: "What is the highest priority experiment the LENR community wants to see conducted?"

    Actually ascoli, I was confusing this. The speadsheets relate to earlier tests?

    As also indicated in my jpeg (1), the only two publicly available spreadsheets refer to the 120 W tests performed in May 2016, whose results were reported in various documents, starting from a JCMNS article issued in 2017 (2).


    In that case have been misapplying them to later tests. I guess I don't know, for R19 or R20, whether the input power is measures with wattmeter, or as V*I.

    It doesn't matter for the moment. Thanks to the availability of the spreadsheets, the only tests which worth to be analyzed are the 120 W tests held in May 2016.

    When JR will release the spreadsheets of the more recent tests (3), it will be interesting to extend this analysis to the R19 or even the R20 reactors.


    Each of these methods has different possible issue.

    The main issue connected with these two methods of measuring the input power is that for the 120 W active test the data directly acquired from the wattmeter have been deleted and substituted by the V/DC*I/DC products. This is a serious methodological issue, rather than a technical one.

    Ascoli65I'm not sure you got the timeline right in terms of when Mizuno switched to internal heater. Jed can confirm.

    Officially, the internal heating was adopted after the May 2016 tests. Internal heaters were mentioned in an article published on JCMNS in 2018 (4).

    But actually the results of the active tests run in 2016, also reported in Figure 4 of the 2018 article, were obtained by inadvertently powering an internal heater, probably the ceramic heater largely described in the 2017 article (2) and also represented in the schematic in Figure 1 of the 2018 article (4).

    (1) Mizuno reports increased excess heat


    (3) Mizuno reports increased excess heat


    In all previous tests with this technique, both the active and control reactors were heated with an external resistance heater.

    OK. So, the correct resistance heater to be powered during this type of test is the external one for both the active and the control reactors. This fact explains very well why Mizuno didn't get excess heat for months and then, during the May 2016 test campaign, he suddenly got a COP of about 2: he simply connected the wrong (internal) resistance heater of the active reactor and at the same time a wrong value of the shunt resistance was manually entered in the data logging and processing system.


    In all cases, with all reactors and with heaters only, ranging from 50 kg down to 300 g, the calibration points are indistinguishable. The data points fall on top of one another within the margin of error. You cannot tell the difference between internal and external heating. You cannot tell the difference between glow discharge and resistance heating. You cannot tell the difference between a 50 kg reactor and a resistance heater alone. It makes no difference what is in the reactor: air, deuterium or a vacuum. The Delta T temperature difference of outlet minus inlet air is a function of the heat produced only. No other factors affect it. It impossible to tell where the heat is coming from.

    This could be true only in stationary conditions. Vice versa, the transient curves, like those shown in Figures 27 and 28 of the Mizuno's paper published in 2017 (1), tell a lot about the position of the heating source!


    This is obvious to anyone who has used a calorimeter. If this is not how it worked, it would not be a calorimeter. The fantasies and nonsense that you and THH post here about magical changes in the way heat affects the air temperature would violate the conservation of energy.

    I wonder how it is possible that calorimetry experts don't understand the elementary notion that the rate at which the air temperature increases reveals the position of the heating source, especially with a reactor whose body weights more than 20 kg. It's just a matter of conservation of energy. When the heater is inside the reactor, the heating of the cooling air is delayed due to the need to heat the reactor body first.


    Your other notions are also preposterous nonsense, such as your claim that a digital power meter can mistake 300 W of electric power going into a resistance heater for 50 W, and for some reason it only does this when there is deuterium gas and a Pd-Ni mesh in the cell. How would it know? How can a power meter be affected by the cell contents? The resistance heater cannot be affected by the cell contents. If it were, I suppose it would break or contaminate the gas.

    You are confusing me with someone else. I'm not interested in the R19 or R20 tests. I've already proposed elsewhere how to deal with such astonishing claims (2).

    Here, I'm only talking about the May 2016 tests, those for which we have a lot more information, including the 2 spreadsheets of the active and control tests at 120 W.


    (2) Team Google wants your opinion: "What is the highest priority experiment the LENR community wants to see conducted?"

    Hi THH,

    thanks for your in-depth lesson on PSUs. Very interesting. Now, I understand your concerns much better, although I have a different idea about what caused the apparent COP of about 2 in the May 2016 tests.

    IMO, a much more straightforward and simple explanation should start by observing that, during the active and control tests, two different resistors were used to heat the two reactors: the active reactor was heated by an internal resistance heater, while the control reactor was heated by an external resistance heater. This fact is evident by observing the trends in the output power curves shown in Figures 27 and 28 of the Mizuno's paper published in 2017 (1). Obviously, this was a serious mistake of the 2016 experimental campaign: it is not possible to use different settings for the active and control reactors. This elementary requirement was known to the tester, but evidently something went wrong during the preparation of the wiring of the experimental set-up.

    The other fact to consider is that the DC voltage and current to the resistance heaters were measured by using two voltmeters and one shunt resistor. Voltmeter readings went directly to the data logger, but the value of the shunt resistance was probably entered manually in either the data logger or the PC for data acquisition at the beginning of each run. Therefore, a simple hypothesis that would explain the error which caused the false COP is that the two resistance heaters (internal and external) were associated with two different shunts, but that the same resistance value (the highest of the two shunts) was used to calculate the value of the DC current, in the false presumption that the same heater was connected both in the active and in the control reactors. If the shunt resistance required for the internal heater was actually half the value of the shunt resistance of the external heater, the DC current calculated (and logged) during the active test was half of the value actually fed into the reactor and this error led to a COP of about 2.

    All the above mess could have happened inadvertently, due to a mistake in wiring the circuitry of the cables coming out of the two reactors.

    On the contrary, what is very difficult to explain is the difference between values reported in the "Input power" column of the active and control spreadsheets. The two tests at 120 W were performed on two consecutive days. So it's certain that in the "Input power" column of both the original spreadsheets there was the same quantity, namely the power measured by the Yokogawa power input analyzer (the wattmeter). This implies that the spreadsheet of the active test has been later modified to remove the original values directly acquired from the wattmeter, by replacing them with the V/DC*I/DC products. This fact would seem to implicitly confirm the above hypothesis about a major, although very probably inadvertent, mistake occurred during the experimental campaign held in May 2016.


    I do not know who wrote that note, so I cannot address the issue. It does not sound like I wrote it.

    For sure, it was not written by Mizuno. He should have known if the "Input power" were obtained by multiplying V and I or came from the wattmeter. Who others could have written that note?


    I am sure there is no significant difference measuring power to a resistance heater with these different instruments. If you disagree, I suggest you try it, rather than spinning out these endless, useless, idiotic fantasies and veiled accusations.

    I don't disagree. Normally, there is little difference, of course. But the crucial question is: why the wattmeter reading was substituted by the VxI product for the active test only? Could you ask Mizuno?


    You and THH live in a fantasy land where nothing is real and anything might be true.

    Well, I can speak only for myself and, yes, after having followed the CF/LENR saga for a few years, I'm fully aware of this!

    Hi Ascoli. I understand that. Implications:

    (1) If heater voltage is DC low ripple than V*I and wattmeter result will be the same (roughly) as they are during the cal run.

    Sorry, THH, maybe I'm not able to explain correctly what I mean, but no, this is not my point. Forget the ripples, please. Assume there is none.

    As you correctly said, V*I and wattmeter results should always be ROUGHLY the same, as shown in the second graph of my jpeg (*), but NOT EXACTLY the same as shown in the first graph of the same jpeg.

    As for the control test, the discrepancy P-VxI - which ranges from -0.4 to +0.4 W - and the difference in the P and VxI shapes assure us that the "Input power" signal came from the wattmeter. Pay attention, please, not from the cheap portable wattmeter (Sanwa WattChecker, 75$) brought with him by JR, but from the expensive wattmeter (Yokogawa PZ 4000, 16000 $) which was connected to the HP data logger.

    On the contrary, the substantial identity between the "Input power" values and the "V/DC*I/DC" product shown in the active test spreadsheet strongly suggests that the original data coming from the Yokogawa wattmeter were substituted, in this only case, by the values obtained by simply multiplying the V/DC and I/DC columns.


    (2) If heater voltage has significant ac component then average V * average I (which you have shown was used) will under-read the real power (as would have been measured by the wattmeter). The underreading could give COP = 2 quite easily, COP = 6 possible but would require a definitely spiky waveform not likely from a bench PSU even when current limiting.

    No, as said before, forget any AC component. A COP of 2 or 6 can only come from a big mess in the input circuitry. IMO the most suspicious candidate to be the cause of these huge errors is the shunt resistor used to measure the DC current.


    (3) I can't say this is deliberate. To manually enter those wattmeter values would be a pain, might only be done once during cal to check V*I is OK. Or, in active runs there might have been need to use a/d inputs for another sensor. It is careless, and bad practice, but not sinister.

    No need to manually enter the wattmeter values. As said before, the Yokogawa PZ 4000 wattmeter is directly connected to the HP data logger. You are making confusion with the portable Sanwa WattChecker wattmeter plugged in by JR.


    PS - OK - you mean perhaps a wattmeter on the INPUT of the PSU!

    Yes, I assume that the Yokogawa PZ 4000 wattmeter is placed upwards of all the circuitry, hence upwards of the PSU too.

    Anyway, I repeat the main question: why these "Input power" values directly measured by the Yokogawa wattmeter and recorded by the HP data logger appear only on the control test spreadsheet and not in the active test one?

    (*) Mizuno reports increased excess heat

    (1) ambiguity on spreadsheet between what is calculated and what is measured is nothing new.

    That wasn't the ambiguity I was referring to. The ambiguity is in the note which appears on the spreadsheet on top of the "Input power" column: "V/DC*I/DC but probably measured directly with a wattmeter"

    The ambiguity is: the data reported in the "Input power" column of the spreadsheets

    were obtained by multiplying the values listed in the V/DC and I/DC columns


    were directly measured with a wattmeter?

    For what I have seen in the two spreadsheets of the 120W active and control tests held on May 2016 (*), the "Input power" of the control test comes from a wattmeter, while the "input power" reported for the active test was obtained by multiplication.

    Can you understand the possible implications of this hypothesis?

    It's not a matter of waveform, rms or ripples! We are dealing with claimed COPs allegedly ranging from 2 to 6 and beyond!

    (*) Mizuno reports increased excess heat

    I don't know how you could add that data, or where it would go, or what use it would be. The spreadsheet data is automatically collected from the HP A/D converter, with additional columns computed. The watt meter does not output data as far as I know. You just look at it from time to time.

    So, I understand that the portable wattmeter you plugged into the wall is a "third check" of the "Input power", the other two being the product V/DC*I/DC and the wired instrument, probably another wattmeter, connected with the HP A/D converter, which provides the data listed in the "Input power" column of the Mizuno's spreadsheets. Can you confirm this?

    If this is the case, the main question is why the measured data from the wired wattmeter, which should have appeared on the spreadsheet of the 120 W active test, were instead substituted by the values obtained by multiplying V/DC and I/DC (*)?

    Can you figure out the possible implications of such a substitution?

    (*) Mizuno reports increased excess heat

    Installed by me. A Sanwa WattChecker, $75.

    Are you meaning that you personally installed this wattmeter in the Mizuno's lab?

    In case, have you done it also for the 120 W tests held on May 2016?


    These things are extremely reliable. The trick is to plug the meter into the wall, then plug the power supply into the meter, and then turn on the power supply without connecting it to anything. That gives you the minimum overhead for the power supply. It will consume somewhat more overhead power and produce more heat with high output, but you can ignore that. It is a reality check.

    In my opinion, expensive instruments should always be double-checked with modern, handheld digital instruments.

    I agree, these instruments provide a simple and reliable way to esteem the upper bound of the electrical power fed into the system.

    But then, why the values from this "reality check" have not been included in the spreadsheet of the 120 W active test held on May 19, 2016 (*)?

    (*) Mizuno reports increased excess heat

    Mizuno’s active and control test at 120 W, May 2016 – Remarks about wattmeter data

    Input power is confirmed with several instruments, including a cheap digital watt meter between the wall and the power supply. There is no possibility of a significant error.

    OK. A digital wattmeter placed between the wall and the power supply is good way to provide an overall estimation of the input power.

    Actually, on the spreadsheets of the active (1) and control (2) tests at 120 W run on May 2016 and just on the right of the "V/DC" and "I/DC" columns, there is a column titled "Input power". An upper note explains better the contents of this column, but provides two possible sources for the data listed therein: "V/DC*I/DC but probably measured directly with a wattmeter".

    In order to solve the ambiguity, I calculated two more values:

    - VxI, ie the product of "V/DC" and "I/DC";

    - P-VxI, ie its difference with respect to the "Power input".

    The next jpeg shows the curves of "Input power", "VxI" and "P-VxI" for both the active and the control tests.


    The graphs reveal a weird mixed situation:

    - as for the active test, which was run on May 19, it seems that the "input power" column contains the product V/DC*I/DC. In fact a large percentage of the value of P-VxI are zeros and the positive and negative spikes can be easily explained as the effect of the rounding off of the numbers;

    - as for the control test, which was run the next day, on May 20, the "input power" column contains for sure the data coming from a different instrument with respect to those used to measure V/DC and I/DC. Very likely this other instrument is the wattmeter you have mentioned.

    Well, as you understand, this situation raises a very critical question: why are the wattmeter readings reported only on the control test spreadsheet and the active test spreadsheet contains the results of the product V/DC*I/DC?

    Can you please provide an explanation or maybe ask Mizuno?

    (1) Mizuno : Publication of kW/COP2 excess heat results

    (2) Mizuno reports increased excess heat

    I think picking apart small details of the calorimetry is silly. Large error sources would be another story but it is puzzling what such a source could be, absent deliberate deception, especially considering the calibrations. I have not read any theories for such errors here. Did I miss it?

    Not exactly a theory, but a clear evidence of a possible source of large errors in Mizuno's experiments (1): the spreadsheets of the active and control tests carried out in May 2016 (120 W in, COP=2 ca.) show the existence of serious inconsistencies about the electrical parameters of the input power. Similar inconsistencies could easily explain the extraordinary results (50 W in, 250 W out) claimed for the R19 tests.

    IMO, before taking into account any small detail of the calorimetry or attempting a replication, it would be necessary to understand what caused the inconsistencies contained in the May 2016 spreadsheets and to carefully look at the data recorded during the R19 active and control tests.

    (1) Mizuno reports increased excess heat

    Many here, I thought, think that classic open or closed cell FPHE D/Pd experiments are well enough understood that excess heat can be reliably found by someone of skill. I'd suggest that that "success reliable" recipe be communicated to google. They can they replicate it, ...

    Yes, exactly, this is what is (almost?) unanimously believed in the CF/LENR community, so this is what should be replicated by Team Google.

    However, you should keep in mind a fundamental difference between open and closed cells. For open cells, we have videos that show the intimate source of the presumed excess heat, ie the sparking glow Pd cathode, and "the dramatic heat effect of the cold fusion reaction on the water fuel". This visible effect can be replicated and unquestionably demonstrated by the Google's researchers. On the contrary, closed cells are sort of magical black boxes equipped with instruments which produce numbers, you should trust into.