Comments on Sur la redéfinition des unités SI : 3. la réalisation du kilogramme jusqu'à 2018

You may need to do slight updates on two of your older writings to acknowledge the changes in SI. <URL: http://www.madore.org/~david/misc/maxwell.xhtml > mentions the definitions of units; <URL: http://www.madore.org/~david/misc/time.html > quotes the then current and earlier definitions of SI units.

Ilia: that is interesting, but it's also one of those experiments that can no longer be performed, because there are no more non-digital TV broadcasts in Hungary anymore since several years ago. (Non-digital radio broadcasts still exist.)

I remember, from a 70's Soviet popular science book that I read as a teenager, a suggestion to determine the order of magnitude of the speed of light by phantom images on the screen of a cathode-ray TV set. The reasoning goes like this: in a CRT TV, each frame is displayed as a series of "scan lines". Knowing the frame rate (apparently 25fps for Soviet TV) and the number of lines per frame (apparently 625), one can determine the time needed for the electron beam to scan a single line: roughly 1s/25*625 = 64 microseconds. In this time, a light ray covers approximately 19 kilometers.

Now imagine that the signal from the TV station reflects on a nearby high-rise building that is a few kilometers away. This induces a radio echo delayed by a couple dozen microseconds compared to the main signal, which leads to a ghost image on the screen shifted by a significant fraction of the screen width.

So if you are able to identify the source of the radio echo (of course this sounds like a big "if"!!!), you can have a rough estimate of the speed of light by measuring by how much the ghost image is offset.

Yes, but the first experiment was before people even knew about electromagnetic waves, and the second was before we had vacuum tubes to create narrow spectrum radio waves. Both the science and the technology has improved a lot since that.

These days, anyone can measure at home that the wavelength of visible light is in the micrometer order of magnitude. You take a cheap transparent plastic ruler with a millimeter scale printed on its back in black paint that is so thick you can feel that the surface of the ruler is uneven as you draw your finger over it, reflect the beam of a cheap red laser pointer device from the ruler such that the beam is almost parallel with the ruler, projecting the reflected light to the wall, and observe that you see a diffraction pattern.

A HAM radio enthusiast can create their own electric circuits that uses transistors to create radio waves of a frequency of his choice. There are a lot of ways to measure the wavelength of radio waves, at least to an order of magnitude. The only thing missing is a way to build a circuit such that you can also tell the order of magintude of the frequency of the radio wave, and if you do both, then you can measure the speed of light.

But before circa 1890, people didn't have vacuum tubes, so I don't think anyone had any way to measure the frequency of radio waves. That's why they had to do experiments that measure the propagation speed of light directly, and those too were much harder to set up without modern technology.

You've explained here how for a century, the definition of the kilogram was based on the PIK, a unique and irreplacable artifact that very few people claim to have seen, and there are no good photos of it. To me, this makes the artifact semi-mithical, and I find it difficult to even believe that it exists and is actually used to calibrate the national kilograms, so the PIK has a status similar to Jesus Christ for me.

However, the new definition of the kilogram accepted today are also semi-mithical in the same sense. The new definitions seem to depend both on the Avogadro constant (the reciprocial of the size of elementary particles), and on Planck's constant. Both of those constants are such that I don't know of a convincing way to prove even their order of magnitude without some very complicated laboratory equipment, so I can only know their value by believing what books tell me. And even with all the fancy equipment that the best laboratories of metrology have, I find it almost impossible to believe that they can obtain eight significant digits.

I haven't yet decided whether the speed of light is also semi-mithical. In the year 2000, I have witnessed an experiment accessible to ordinary mortals that seemed to establish the order of magnitude of the speed of light. Or so I thought at that time. But that experiment actually seems to prove that the speed of light is around 3*10**7 m/s, ten times slower than the actual value. So I now believe that either the experiment was theoretically flawed, and was measuring an unrelated effect, or my memory of it is significantly flawed. Alas that experiment is impossible to reproduce now. Some people argue that we can't be mistaken about the speed of light, because our GPS receivers work. But all that proves is that the engineers making the GPS receivers know the speed of light. If the speed of light was actually 3*10**10 m/s, but my textbooks mistakenly told me that it was 3*10**8 m/s, then there would be no experiment I could perform to disprove what the textbooks say, and GPS receivers would work all the same. I wonder if there's at least an experiment to measure the speed of light that a HAM enthusiast can realistically perform.

It appears that the vote has happened now, and officially accepted the new definitions. The votes were public, and apparently the delegates of the member countries were asked in alphabetical order of the French name of the country. Ruxor <URL: http://www.madore.org/~david/weblog/d.2012-08-07.2061.html > will find that the secre2ary general of the CIPM pronounced <URL: https://youtu.be/qA67T7FPBME?t=8870 > the French name “Israël” with an [s], but the English name “Israel” with a [z].

Bonjour
«l'air est environ 2×10−4 fois plus léger que le platine iridié».
Excusez-moi de vous demander pardon, mais ça me choque un peut cette phrase. On a plutôt l'impression que l'air est 1/(2×10−4) = 5000 fois plus léger que le Pt iridié.
Autre chose : vous écrivez « De nos jours, je crois comprendre que toutes les manipulations sur la balance sont faites de façon automatisée ». Si c’est le cas, c’est très bien, car tout le monde procède ainsi de la même manière. Mais il me reste un doute : qui (ou quel organisme) calibre la balance ? En gros, pour calibrer la balance, il faudrait un étalon de travail, or si c’est justement celui sur lequel on réalise la mesure de la masse, le truc devient complétement circulaire… Le fait d’étalonner au moins deux masses (A et B dans le texte) est-il suffisant pour écarter cette ambiguïté ?

Right @Ruxor, j'avais commencé à faire rapido un excel en cm et je suis passé en mm mais n'ai pas forcément bien remis à jour tous les multiples - et les ordres de grandeurs étant assez inhabituels, cela ne m'a pas interpellé finalement :-(

reprenons rapidement :
Griffe de 5 cm de long, sur 0.1 mm de large et 0.001 mm de profondeur
volume = 5. 10^-12 m^3
soit (21,45 g/cm^3) => 107 microgrammes

reste que j'ai quand même du mal à le visualiser. J'ai un microscope 500x à la maison, je vais faire un test en essayant de compter les pixels avec qq règles de 3 :-)

passionnant et bien rédigé

je me suis demandé quel poids pourrait être perdu sur une griffe de 0.1 mm de long et sur une profondeur (en biais) de 0.001 mm (ce que peut mesurer un profondimètre comme MICROMÈTRE DE PROFONDEUR DIGIMATIC MITUTOYO 329-250-10) de platine.
en gros, j'obtiens … 1070 microgrammes. C'est fou…

Whaou….je n'avais jamais tilté sur le fait que ces mesures sont faites dans l'air. C'est assez fou d'un point de vue théorique mais ça se comprend historiquement.

Dis-moi si pour obtenir un étalon des poids et mesures on a tant de mal alors pour obtenir une pensée politique universelle ça doit être coton !
…
Enfin les physiciens sont des gens presqu'aussi obstinés que les mathématiciens - il y a donc de l'espoir pour un jour prochain et un avenir radieux du système SI.

J'ai commencé à feuilleter en diagonale, et ça s'est révélé tout à fait passionnant même pour un lecteur à la recherche d'infotainment plutôt que d'information scientifique.

J'ai été complètement épaté par l'apparition des chamois. C'est à la fois heureux et la fin d'une époque de penser qu'on n'aura plus besoin de chasser le chamois pour définir le kilogramme.

Autrement, fascination pour l'histoire de l'étalon "donné à la République Tchèque". Que doit faire un État pour se faire offrir un étalon secondaire ??? Plus sérieusement (mais pas vraiment non) sais-tu si tous les signataires de la convention du mètre ont un étalon national ou si c'est en pratique l'apanage des pays ayant un minimum de poids scientifique ?

Je trouve ça fou qu'il soit si difficile de peser un kilogramme à 50 microgrammes près. 50 microgrammes, ce n'est pas du tout une masse microscopique : c'est un cinquième de grain de pavot. Un grain d'acier, et même de platine-iridium, de cette masse est sans problème visible à l'oeil nu ! J'imagine qu'une rayure correspondant à la disparition d'une telle quantité serait facilement visible aussi…