# New measurements of kilogram

1. Oct 16, 2015

### exponent137

What is the point of new measurements of kg, by $Si^{28}$ ball and with Watt balance? Is it this a special moment because Avogadro numbers, $N_A$, of both measurements agree? What values of $N_A$ were given by both measurements and what values of $N_A$ were before these measurements? I found some links but this aspect was not specified.

2. Oct 17, 2015

### exponent137

I obtained an answer from jfizzix by a mail report, but here it is not visible. (Maybe some software bug.) He gave a good answer. Beside this, I am still interested in a "play of numbers", which numbers now agreed from Si-28 and Watt balance and what these numbers were before this agreement, this means that they were different before.
Watt balance fixes Planck's constant and Avogadro project fixes Avogadro number, as jfizzix said. I suppose that Watt project gave either different Avogadro number, or different kg, before this last measurement? But it is known in the last measurement they finally agree?

Last edited: Oct 17, 2015
3. Oct 17, 2015

### Staff: Mentor

jfizzix deleted his message. There is no way the forum can cancel notification mails after they have been sent.

The numbers should be in the publications.

We can fix both the Planck constant and the Avogradro number, then we have to give up the definition of 12C having a mass of 12 u. I'm not sure if that would be useful. Masses of atoms relative to 12C are much easier to measure than masses of atoms relative to a kilogram.

4. Oct 17, 2015

### exponent137

Does this means that Avogadro project calculates 12C and compares it with an old measured 12C? But either Watt balance uses this Avogadro number and then also calculates 12C. Or both measurements together calculate 12C and compares it with a value known before, independent of these two measurements?

This my question is like, what is number of unkonwns and what are equations for these unknowns.

5. Oct 17, 2015

### Staff: Mentor

It cannot.
It is trying to express the kilogram as "N atoms of 28Si", which can be translated to "M atoms of 12C" using the relative masses of those atoms. You can compare the new number N (or M) to older measurements, the idea is just to get more precise (and more precise than the kilogram prototype in Paris). If the measurement is precise enough, "N atoms of 28Si" could become the new definition of a kilogram.

The Watt balance is unrelated to atomic masses. It tries to measure the Planck constant (in SI units) more precise than before. Precise enough so we could fix it to a specific number, and use this as definition of the kilogram in the future.

6. Oct 17, 2015

### exponent137

Thus, we have one "definition" of kg from 28Si sphere, and another "definition" of kg from Watt balance. (I write "definition", because they are not yet offical.)
Does this means that these both definitions of kg now agree inside of measurement error, but before they disagred, because their measurement errors were smaller than their difference? Or, the second option: Before one of these measurements was not precise more than prototype kilogram IPK? I think that the Watt ballance was this problem?
When new measurement will be more precise, disagreement can happen again?

7. Oct 17, 2015

### Staff: Mentor

There is nothing they could agree or disagree on. They measure different things.

8. Oct 18, 2015

### exponent137

http://arxiv.org/pdf/1409.7597.pdf
https://en.wikipedia.org/wiki/Proposed_redefinition_of_SI_base_units
It is also important that Avogadro project measures $N_A h$, not only $N_A$.

I think that the last achivement was:
"In August 2015, when CODATA published its latest value for Planck’s constant, the uncertainty was 12 parts per billion, just over one-quarter of its value in CODATA’s previous report — and within the CIPM’s requirements."
http://www.nature.com/news/kilogram-conflict-resolved-at-last-1.18550

What were CIPM's requirements?

p.s. I think that this is big step forward, because macroscopis bodies will no more be references for units. It is said that some measuurement will became 50 times more precise because of this reason.

Last edited: Oct 18, 2015
9. Oct 18, 2015

### Staff: Mentor

The disagreement seems to have been within the Watt balance method, see this article:
An accurary better than the comparisons of the kilogram prototypes, but I don't find the precise numbers.

10. Oct 18, 2015

### exponent137

Maybe something connected with the second figure, because of time drift of the prototypes masses?

Probably, despite of drift, preciseness of measured masses is better than scattering of results, because the measurement of drift is possible.

Last edited: Oct 18, 2015
11. Oct 19, 2015

### f95toli

Not quite true. The goal of the experiments on the Si28 is actually also to measure Planck's constant (albeit indirectly); it is NOT a new way to realize the kilogram.

The reason is simply that in the new SI the kilogram will be defined via Planck's constant, meaning the latter will be given an exact value, similar to what was done to the meter when c became a constant.

Now that the Si28 spheres and two Watt balances agree on the value of h, the likelihood if this change being made in new SI in 2018 has increased.

Another interesting consequence of h being defined is that whereas the new ampere will probably be defined by giving e (I=e*f)an exact value we will now also be able to realize the Ampere by combining the the Josephson effect (the Josephson constant h/2e will now have an exact value) and the quantum Hall effect. This is quite important since we do not yet have electron pumps or counters that are accurate enough to actually emit/count electrons with the required precision (ideally better than about one part in 10^8). Both the Josephson voltage and the Hall effect can be measured with extremely high precision.
This is in fact how the Ampere has been realized for the past 25 years or so, but so far it has been "unofficially" (using a CODATA value for the Josephson constant).

12. Oct 20, 2015

### exponent137

What is still possible to improve at Si28 sphere? Maybe to reduce number of defects in crystal structure, to reduce number of isotops, to improve roundness of ball? Or is everything here possible, but measurement of $h$ with Watt balance is now a vital factor?

What are other elements or molecules, which can be useful for clear crystals?

How large pure crystal of any matter is maximally possible to make? I think for measurements of gravitational constant.

13. Oct 20, 2015

### Staff: Mentor

Everything can be improved, the roundness got significant attention but I don't know if it is the limiting factor in precision.

Silicon has the infrastructure for large and high-quality crystals from the semiconductor industry.

Measurements of the gravitational constant don't need special materials - measurements of the mass and shape of the object are much more precise than other issues.