TD replications

[Floor]

@Gotoluc

Great work Luc !

I am setting up to do measurements / replication on you last demo.
This will be of an output stroke only.

If... I also do the input stroke measurements, they will have to be
of a straight on (parallel) stroke, and one magnet only (input)....
to one magnet only (output)

@all readers

These experiments are being done on the fly. (not perfect)
My measurement process will be only an approximation of
Luc's process (not identical).

please find the attached PNG file

                   regards

[Nonlinear]

Thanks Luc for the video, now your results make more sense.
There are several mistakes in your approach and measurements, which may be very well the cause for the detection of COP>1. The biggest mistake is to judge the efficiency of a device based on average forces. That is completely unscientific, and it can very easily mislead you. The total work must be measured and calculated, like in my spreadsheet.

The second mistake is ignoring the role of measurement errors. A measurement is never 100% accurate, there is always some error in it. The experimenter must be aware of the expected maximum error margin of his measurements, and disclose it together with the measurement results. Without this, the data can not be taken seriously. For example the best resolution of your scale is 5g which is extremely low and produces a very large error if you are measuring forces in the range of 0 to 100g. When you are measuring 100g then the uncertainty of the measured result is 10g, which is 10%. The correct way of showing your measurement result is: 100g +-5g, or with other words, the real force could be anything from 100-5=95g to 100+5=105g, the error margin is 105-95=10g, which is 10% of the measured value of 100g. Therefore if you find a COP=1.1 with such large error margin of measurement, then your measurements are pretty much useless to prove anything. If you are measuring even less than 100g, like in some of your measurement series measuring 5, 10, 20g etc. then your error margin is so huge that the data is of no value to prove anything. If your scale has a low resolution, then build a device that requires the measurement of about 100 times larger forces than the 5g resolution. If this is not practical, then use a scale that has got sufficient resolution and accuracy to produce around 1% (or less) measurement error.

The third mistake is not to measure the complete cycle of movement. For example webby1 was trying to convince you few pages back that you have to measure the 4th part of the cycle as well, in one of your earlier devices. He finally succeeded in this effort in post:
http://overunity.com/16987/td-replications/msg496971/#msg496971 and you provided the data in:
http://overunity.com/16987/td-replications/msg496974/#msg496974
After you have measured this 4th part of the cycle and taken it into account, then your (still incorrect) averaging calculations showed only 10% of excess work, which can very easily attributed to the other mistakes already mentioned.

In this case of rotary device this is not as critical as in the previous devices, but I would still recommend you to measure the complete cycle. Meaning, two rotary magnets pass in sequence above the slider. Please also measure the output force as well at least 10 times, like after each 1mm movement.

I was trying to get a manual feel of the forces in your device using two neodymium magnets of 4x2x1cm, which I don't recommend to anyone. These magnets are just too powerful, and if one doesn't have very strong fingers, they can also harm you! But, I have got no ceramic magnets of rectangular shape right now, so can not do the safe version.

Anyway, keep up the good work, and if the COP is still higher than say 1.2 even after fixing these mistakes, then it should be possible to build at least a perpetuum mobile using this magnet arrangement. If the COP would be really 1.6 like in your measurements, then the machine should be able to generate useful output power as well, besides just running itself.

Thanks Floor for the links and the drawing.

[gotoluc]

Here is a sneak peek of the v2.0 Magnet Torque Amplifier device before it's all assembled and unable to see the internal design mechanism.

Link to video: https://www.youtube.com/watch?v=BMVES42VbzA

Test results should come in the next couple of days

Stay tuned

Luc

[gotoluc]

Here is the device assembled


Link to video: https://www.youtube.com/watch?v=OsEbX8yJ91I


I'll need to bolt this down for testing


Luc

[Floor]

@All  readers

The below png file contains the results of  a
quasi replication of GotoLuc's MagTorqAmp

    (output only)

             best wishes
                   floor