Naked Science Forum
Non Life Sciences => Physics, Astronomy & Cosmology => Topic started by: Mike Collins on 13/05/2008 08:41:43
-
Mike Collins asked the Naked Scientists:
Hello Chris,
I have a question about how electrical contacts work, such as in a switch.
We are taught that the phenomenon of current is the drift of loosely coupled, outer-shell electrons from one atom to another. I would have expected that in order for this to happen, the atoms involved must all
be part of a continuous solid of the conductor. However, if two pieces of copper are brought together, current flows across the junction, so they actually come close enough for electrons to jump from the atoms of one to those of the other. When they are separated, however, they maintain their original surface profile, so my question is this : What is going on in the boundary layer of atoms that they allow the drift of electrons from outside, yet they don't fuse together into a continuous solid? The atoms of each piece of copper stay with their original body.
In pondering this, I thought that some atoms may be transferred from one side to the other, so maybe the two bodies fuse together, but on such a small area that they are so easily broken apart that the force required is unnoticeable, but on further thought, I think that the pitting that we see on contacts is due to material that is transferred in the arc that forms as they are separated. If, however, the current is stopped before the two contacts are separated, I don't think there will be any transfer of material. Am I wrong?
Another aspect about this that I find intriguing is that if dissimilar conductors are brought together (eg copper and silver), a current is conducted, yet the silver atoms stay on one side and the copper atoms on the other.
This question has intrigued me for so many years that I'm really hoping it can be aired (pun intended!)
Best regards,
Mike Collins
Basingstoke,
UK.
What do you think?
-
Like you said its the loosely coupled outer-shell electrons that are moving, not the closer tightly bound ones that are holding the atom to the rest of the atoms in its structure. The actual atoms of the conductor aren't travelling anywhere, the electrons flow around them.
-
It is an interesting question. The two surfaces will be very rough on a microscopic level so I suspect we would just get multiple point contact areas. These would allow electrons to move across from one surface to the other. In practice surfaces do need to be clean and pressed together to get a low resistance contact. With copper there can also be an oxide barrier to overcome. I think that it is also the case that if there is a gap (either an air gap or one caused by a thin layer of oxide) that this can easily be broken down by the field that would result from the initial potential difference. During this period electrons would flow across the gaps from the higher field pointy bits on each side. Enough flux could cause local heating and microplasmas which may result in some material movement too.
As I said, an interesting question that deserves some thought.
-
It is a wrong definition of electric current like a charge movement.
Electric current is is magnetic moment perturbation along conductor. But in order to get the correct answer a new definition of metallic bound must be provided (it is already on the net). Look a Little bit at Trouton Noble experiment and relativity post on this forum and You will see another face of physics.
Regards
Sorin Cosofret
-
One problem with mechanical switches is that they also suffer from "contact bounce" for a few milliseconds after switching and then settle down to a switched state.
The switch does not close cleanly no matter what the size and happens so fast that with most circuits you will not notice it. Its on a microscopic level so I wont explain it in any detail as it would take too long so give it a look up as to why.
Its not very noticeable until you use a switch to trigger or switch a logic circuit. Because of the fast time response of digital circuitry by closing a mechanical switch you can get multiple triggers because of the bounce. In that situation you have to use a switch "de bounce" circuit so you only get one clean pulse or trigger when the switch contacts close.
Although, now you can get "Quantum tunnelling pills", so something else to look up. [;D]
-
You can get a clean contact switching by using a mercury wetted relay.
-
Why is this cleaner, Graham, and how does it work?
-
One problem is inductive loads, this causes a lot of arcing on the contacts. A car ignition system is a good example of this situation and they have a capacitor across the contacts to minimise the arcing.
How do you keep the mercury on the contacts?
-
The arcing is much more severe when breaking the circuit, when there is already current flowing. The induced emf is proportional to the rate of change of current times the inductance and breaking the circuit calls for a fast current change. A reverse biased diode is often placed across inductances such as relay coils and this damps any resonance, killing the spikes.
-
(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fhome.att.net%2F%7ETom.Horsley%2Fphonetale%2Frelay.jpg&hash=002c53c341e9aca55c7e83b476e723db)
The diode is across the coil of the relay to stop the back emf doing any damage.
(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fwww.worldproducts.com%2Fimages%2FTable2A.gif&hash=57808bd6182a8f1c15c798728a4f6e60)
A capacitor goes across the contacts to stop arcing.
-
Although a capacitor across the contacts helps when the contacts open there is a large current when the contacts close short circuiting this capacitor which is now charged to the full system voltage.
To avoided the resultant damage to the contacts and the resulting electromagnetic radiation a resistor should be included in the circuit.
I am glad to say car ignition systems have long since ceased to use mechanical contact breakers.
-
You drive one of these new fangled things, do you?
-
Yep, with 6 little ignition coils driven by multi ENIAC complexity electronics.
-
Cool!
And my boat's diesel engine will run (if you are fit enough to turn it over) with no electrics at all. 34yrs old and still going strong.
-
Turbojet engines don't need electrickery to run either, although it's pretty essential for monitoring.
-
I think the real small ones for model aircraft run computerless but full size ones could not work without electronic control.