Measurement Error and LENR: Why scaling up small Cold Fusion claims never works

  • For total efficiency you multiply that by the electromechanical efficiency of the motor, which for small motors is small unless they go very very fast


    It depends on the type. Brushless DC fan motors are smaller, lighter, quieter, cooler-running, more durable and up to 70% more efficient than conventional AC fans (say 15 watts instead of 50 watts for a fairly powerful fan). However, there are some losses in the ESC systems that control them, where used.

  • Are you talking about motor+fan systems, or just "electromechanical efficiency of the motor" which is what your quote seems to suggest? Fans I know little about, motors I know too much about.


    Both, since I suppose fan efficiency can be quite high. In the mouser quote the 40% efficiency is I think total, not just motor - since it talks about high fan efficiency.

  • While they document lots of maybe not significant "typo" errors, there are very significant ones as well. They do a good job of explaining why the control and active readings can be different. They also show the active flow rate wrong by 100%.

    In the rewrite it seems M acknowledges the 100% error in flowrate, but does not change his power results?

    I have been out of touch, and I missed this message.


    Mizuno did change the power results. The early drafts of the paper showed one level of heat based on the wrong orifice size. This was later corrected. I have all of the drafts in English and Japanese and I am sure of this. The power levels were corrected. However, the size of the orifice was not changed for a while. So, looking at different preprint versions of the paper you may have the impression that the orifice size changed abruptly but the reported power levels did not. The problem was that the two were out of sync.


    That's partly my fault. I looked for problems in wording and language, but I ignored the equations and the arithmetic. Bocijn and others looked closely at the spreadsheets, equations and arithmetic and made many changes to the manuscript. I hope the final version in the journal is correct.


    In the past, in other papers, I have seen large errors survive version after version, and peer-review after review. In one recent case the title was wrong! It was for another paper. It is surprising how people (including me) can overlook large errors. In the 1930s, a British historian was writing a book about one of the King Edwards of England. I don't recall which Edward, but it was #1 through #7. Anyway, the book got all the way through galley proofs and into print before someone noticed it said "Edward VIII." Everyone overlooked it because #8 and Mrs. Simpson was on everyone's mind.


    If this sort thing bothers THHuxleynew, he should refrain from reading preprints and manuscripts. They often include serious errors. I have manuscripts from famous scientists with errors as bad as this. Mizuno tends to be sloppy at times, but even methodical, detail-oriented people made mistakes. Perhaps I should refrain from uploading preprints, because it invites attacks from people who read only the first versions and not the final version. I find it is useful to upload them because some people such as Bocijn help find errors.


    There are problems like this in other fields. In my opinion, Google programmers are among the best in the business. Their software has fewer bugs than, say, IBM software did in 1979, or MIcrosoft does today. Yet I have discovered fairly obvious problems such as the PDF implementation internal paging starting at 0 instead of 1. What impresses me is how few of these errors they make and how quickly they acknowledged and correct the problems when I pointed them out. If THHuxleynew is looking for people who never make mistakes in the first place, he will search in vain. He can hope to find people who acknowledge and correct mistakes.

  • JedRothwell " I have manuscripts from famous scientists with errors as bad as this."


    I am not famous or even infamous but

    in one submitted paper I made the silliest error of writing "Wilcoxon Signed Rank test " instead of "Wilcoxon Rank sum test"

    This was unnoticed/unchecked by my two coauthors who were concentrating on the language rather than the math


    It was a major reason for rejection. Fortunately we were accepted by another journal with twice the impact factor.