# Defining the Ampere

1. Oct 28, 2013

### lionely

1. The problem statement, all variables and given/known data

This may be a stupid question but... I don't understand why the length of the wire has to be infinite in defining the ampere.... could someone explain this to me?

2. Oct 28, 2013

### phinds

Uh ... huh? Where did you get that definition?

3. Oct 28, 2013

### lionely

My text book says... One ampere is that steady current which when flowing in each of two infinitely long, straight parallel wires of negligible cross-section placed one metre apart in a vacuum causes each wire to exert a force of
2 x10^-7 newton on each metre length of the other wire.

4. Oct 28, 2013

### phinds

Weird. I've never seen that one before. An amp is simply a flow of one coulomb per second. Are you quoting this exactly? That is, are you sure that it is presented as "THE" definition of the amp, or as simply a result of the definition of the amp?

5. Oct 28, 2013

### Staff: Mentor

The definition of the ampere has changed over time. See, for example, this entry at NIST.

6. Oct 28, 2013

### Staff: Mentor

That makes the "end effects" vanishingly small. You can ignore them because they are not there.

So here they are defining the Ampere using the relation: F = B i l
where B = μ H

http://physicsforums.bernhardtmediall.netdna-cdn.com/images/icons/icon6.gif [Broken]

Last edited by a moderator: May 6, 2017
7. Oct 28, 2013

### lionely

Oh I see, so it is just to ignore the force on the end of the wires...

8. Oct 29, 2013

### Staff: Mentor

Those were the good times, but these pesky metrologists have to complicate everything.

9. Oct 29, 2013

### D H

Staff Emeritus
That's been the definition for a long, long time. Not an infinitely long time, but quite long nonetheless. The end effects are getting vanishingly small. (You have to be over 80 to remember the old definition.) The ampere currently is a base unit in the metric system and the coulomb is a derived unit, an ampere-second.

10. Oct 29, 2013

### f95toli

The definition WILL change to "flow of X number of electrons per second" in a few years, but current electron pumps (as the devices are called) are not quite good enouh to replace the present definition; the pump accuracy needs to be improved by about a factor of 5-10 or so.
This change will happen at the same time as defining the value of e (and perhaps a few more constants, e.g. h) to be exact (similar to what has been done for c)

11. Oct 29, 2013

### D H

Staff Emeritus
That's right. There's an ongoing effort to completely overhaul the International System of units (SI).

The ampere is currently defined the way it is because of those "pesky metrologists." Measuring charge is a touchy business. Measuring current is a better way to go. The standard presently has the ampere as a base unit and the coulomb as a derived unit. The reason is that experiments that measure electromagnetic induction currently provide significantly greater accuracy and precision than do experiments that measure the electrostatic force.

The movement that is underfoot is to completely rectify the quantum and relativistic scale worlds with our everyday macroscopic world. One consequence of this is that current will be defined in terms of a flow of a number of electrons per second. The ampere will be a "base unit" in name only. When the CIPM is finished with the overhaul, the only true base unit will be the second. All other units will flow from the second based on some defined physical constants.

12. Oct 29, 2013

### phinds

Bolded part is not true. I took EE 50 years ago and coulomb/sec was the definition then.

13. Oct 29, 2013

### D H

Staff Emeritus
No, it wasn't. Fifty years ago was 1963. The current definition of the ampere and coulomb had already been in effect for 15 years. Your teacher at that time was old enough to remember the old definition and wasn't playing attention in 1948 when the ampere and coulomb were redefined.

14. Oct 29, 2013

### Staff: Mentor

NIST page linked to earlier in the thread states the definition changed in 1948. I was taught C/s somewhere in mid seventies (1975, 6?), and I don't think my teacher was old enough to remember the old definition, she was in her thirties or forties.

My bet is C/s was easier to swallow for 13/14 years old than two infinite wires attracting each other with 10-7N per meter.

15. Oct 29, 2013

### D H

Staff Emeritus
Not just to 13/14 year olds, but also to the teachers of 13/14 year olds. I don't know about Poland, but here in the US, primary school science teachers do not have to take the classical electricity and magnetism (E&M) class that vexes so many students of physics.

16. Oct 29, 2013

### phinds

OK, I'll give you that. He was an OLD guy and never was able to understand transistors.

17. Oct 29, 2013

### D H

Staff Emeritus
I suspect that many instructors (and textbooks) still teach amperes as being coulombs/second. My very old Halliday & Resnick (1967) does not. It explains why current rather charge was chosen as the base unit "for practical reasons having to do with the accuracy of measurements."

On a conceptual level, that charge is basic makes a lot more sense than current being basic. However, it's the other way around on the basis of what is more observable experimentally. Currently. Sorry for the bad puns.

18. Oct 29, 2013

### f95toli

It is perhaps worth mentioning that no one actually uses the definition of the Ampere, the realisation is far too complicated even for metrology labs so no one has done the actual experiment in many years (80s).
For the past 20 years or so devices using the Ampere (Ammeters etc) have been calibrated "indirectly" by creating a known current using the realisations of the Volt (Josephson voltage) and Ohm (quantum hall effect). The reason for this is that we can measure these with much, much higher accuracy (we can realise the volt to one part in 10^15 since it relies on the second )than the Ampere in the "real" realisation, another reason is simply that most NMIs maintain system that can realise the Volt and Ohm anyway.

This is of course not an "official" way of doing it since one needs values for the Josephson and Klitzing constant which since they are not defined are not exact; but the CODATA values are so good that this is not really a practical problem. Hence, if you have a calibration certificate for you multimeter this is what the current specs are ultimately refered to (via some intermediate secondary standards)
That said, this is of course not a satisfactory situation which is why so much effort is going into finding a practical method for realising the ampere that is somehow connected to time*.

*Note that this might NOT turn out to be an electron pump that "counts" electrons; an alternative method would be to use a quantum phase slip junction which is the "current analogue" to the Josephson junction (used for voltage calibration).
A QPS junction should when irradiated by microwaves generate current steps (again analogoues to the Shapiro voltage steps in a Josephson junction) that could be used to realise the Ampere with extremely high precision; that is assuming we can get the steps flat enough and the current magnitude is high enough to be useful. This is very much work in progress.

Last edited: Oct 29, 2013
19. Oct 29, 2013

### Staff: Mentor

C/s definition has one important advantage, especially in chemistry - it shows the link between reactions in redox systems (where electrons are transferred between atoms/ions/molecules) and the current. Quite often I see students that don't see the link and I often wonder if it is not because they are not aware of the fact current is nothing else but charge transfer.

But perhaps that's not a matter of definition used, but just of the general lack of knowledge and inability to connect the dots to see a picture.

20. Oct 29, 2013

### epenguin

The definition is going to change next year.

IMHO :

Last edited: Oct 29, 2013