What limits the current in wires

[Kokoriantz]

What limits the current in wires?
I designed numerous transformers obeying the instructions to look up on AWG table, the thickness of the wire for the max RMS current it should carry. The AWG  table has two rows of maximum current, transmission and chassis wiring. I always wondered what is chassis current, until a phenomena presented. I designed a great number of class A amplifiers and in recent times, I made one out of a chip with two bridged outputs paralleled and biased inversely.When I fired the amp, the bass tone was very weak. I paralleled the output thin  pins with thick wires and the bass resuscitated. I wondered how the capacitors do carry high pulse currents? The leads are magnetic, not pure copper, for ceramics or electrolytics. I also found out that old metal case transistors, all had magnetic leads.
I understood that chassis in AWG is steal chassis that constrains the magnetic field of the wire deeper in it as the magnetic wires do. The new transistors, instead of using magnetic leads, they have C & E leads parallel so that their opposite magnetic fields cancel each other. With chips, the +V, OUT,-V are aligned as the output current is always inverse to either supplies and get their magnetic field repelled and constrained. By this, the thin lead of chips 0.35mm × 1mm equal to AWG between 22/23, has maximum transmission current of  0.8A becomes 6A.
For the last chapter of my life, I will deal with physics science rather than engineering, trying to understand where the capacitive and inductive energy in vacuum is stored. It has to do with a deformation type of Einstein's space-time fabric other than the gravity. This current limit phenomena might open another gap, maybe I can see through it  another aspect. I want to understand how the magnetic field effects the electron flow limit.
My son, MSEE, PhD in nanophysics, has never heard about this subject, but many electrical engineers who design the above mentioned component know about it.
If you have any idea, even lunatic one, can always inspire.           
 

[IanB]

What limits the current in wires?

Heat, basically. When the wires get too hot they melt, and then the current stops. If you're unlucky, the heat starts a fire before the wires melt, and then extensive damage can occur.

[inse]

Above that, inacceptable voltage drop, especcially on longer wires, may also be a limiting factor.

[Kokoriantz]

What limits the current in wires?

Heat, basically. When the wires get too hot they melt, and then the current stops. If you're unlucky, the heat starts a fire before the wires melt, and then extensive damage can occur.

One may agree that a wire gets cooled by the chassis and carries more current if it runs upon it. How about two identical wires, carrying the same current of opposite polarity can carry much more current if paralleled than individually, the heat will be higher as less cooled by air.

Above that, inacceptable voltage drop, especcially on longer wires, may also be a limiting factor.

There is no doubt the electric field intervening in this story. By having two parallel wires carrying opposite phase currents get their electric fields cancelled at high frequency. How about steal content capacitor leads, they have higher resistivity than pure copper hence higher voltage drop but can carry much higher currents.

[IanB]

The concept of "current in a wire" is only really true for DC. At DC, it makes no difference whether the wire is alone, or is next to another wire, or is next to something else. The only consideration at DC is the size of the wire and cooling from the surroundings.

At AC the calculations and mathematics become much more complicated, especially at high frequencies. At AC the current does not exactly flow "in" the wire, it flows "with" the wire, and the appropriate description involves transmission line models. However, with transmission lines the behavior depends on the shape of the wire and on the surroundings, such as parallel wires, ground planes or shields.

[mariush]

I would say how much voltage drop your circuit is willing to accept, and how much heating they can handle.

Due to wire's resistance, the more current you have the higher the heat - there's gonna be a rating for amount of heat above ambient if the wire is by itself, and a different rating if the wire is in a bundle of wires (for example 3 wires that are wrapped in another insulation / sleeve) - a wire on its own will be cooled a tiny amount just by air movement around it. When you have a bundle of wires there's no air flowing between the wires.

As for voltage drop ... it's basic formula V=IxR ... the thinner the wire, the higher the resistance and the more voltage drop. for example AWG18 wire has around 0.021 ohm per meter ... so at 10A of current, you'd have 0.2v drop on one meter of wire.

[Kleinstein]

Normally the heat is the limiting factor. With very strong electromagnets the magenetic force can also become an issue and mechanical can tear appart the wire, especially solft copper.

The steel core capacitor wires can not really handle more current. The wire is usually quite short and capacitors may deliver high peak current. For a short time any wire can carry more current.

2 wires close by can not carry a higher current per wire, but more like a reduced current as the 2 wires can interfere with heat loss. 2 wires in parallel may still have better cooling than 1 wire of twice the cross section, but it depends (e.g. does not apply for a transformer).

[SiliconWizard]

And of course at AC the skin effect adds to this.

[Siwastaja]

I = U / R

Assuming constant voltage drive, what happens to current if you increase R by making the wire thinner and/or longer?

[pardo-bsso]

Kokoriantz,

Could you also show us the back of that pcb?

And, leaving aside your other comment about the physical phenomena that might be of importance at higher frequencies and power levels, if you build another amplifier with the same construction does it show the bass issue before adding the other wire in parallel? (asking to try and see if you did more than one thing and one of them fixed the problem)

[Kokoriantz]

You find bellow another AWG table giving the fuse current list. Note that it is independent from the length hence the resistance of the wire. The first who got the idea of fuse is not any electrical expert but a French watch maker A-L Breguet. He was commissioned to produce precision components for telegraphy, and after a lightning disaster which distroyed kilometers of wire, he proposed to install a thinner short wires. Although the resistance of these wires are negligible in series with the intercity telegraph one, it protected it.
I don't know who did realize that the magnetic field is a parameter of the current limit. I suspect Edison and his team to be as he did patent the Fuse.
You see bellow a picture of some components that have magnetic leads, the TO92 transistor is 2N3906.   

[Kokoriantz]

With this amplifier in class A, there is about 1A dc current flowing through it and no external feedback. The LM3886, notorious to lack bass although feedback, has the pins, +V, NC, Out, -V.
The +V and Out current passing through 0.35×0.95 mm thick and away from each other, have low limit current. The remedy is to link the +V to NC at the root of the IC by a magnetic lead of capacitor, and power only from NC, the bass gets greatly enhanced.       

[Kokoriantz]

Normally the heat is the limiting factor. With very strong electromagnets the magenetic force can also become an issue and mechanical can tear appart the wire, especially solft copper.

The steel core capacitor wires can not really handle more current. The wire is usually quite short and capacitors may deliver high peak current. For a short time any wire can carry more current.

2 wires close by can not carry a higher current per wire, but more like a reduced current as the 2 wires can interfere with heat loss. 2 wires in parallel may still have better cooling than 1 wire of twice the cross section, but it depends (e.g. does not apply for a transformer).

The two wires carrying opposite phase currents, as a supply and an output. The heat of two wires doesn't decrease if they are paralleled but increases as individually have better air surrounding.
The magnetic field is not any external but the field of it's own current. If this field is constrained in the wire, the bottleneck gets wider. My question is WHY? The AWG table shows the increase of current limit for chassis wiring is only a factor of 1.2 for very thick wire of 11.6mm diameter, gets up to 7.7 for about number 20 to fall to 6.5 for the last number.

[Vovk_Z]

I would say how much voltage drop your circuit is willing to accept, and how much heating they can handle.

I guess TS wants to discuss a root physical problem - WHY current heats wire and why it causes a voltage drop. So, I guess he rather wanted to ask 'How to make a superconducting wire at room temperature?'.

[ejeffrey]

In most AWG wire gauge charts, chassis just means "not in a bundle of other heat producing wires".  Other factors affecting ampacity is the insulation and termination heat resistance, and whether it is in a metal conduit.  These charts are exclusively for 60 Hz AC or DC and don't consider skin effect, proximity effect, or other magnetic field issues.  It's just thermal.

These charts are also for relatively long wires.
 Short wires, including what might be considered "chassis wiring" by many people doing small electronics are mostly cooled through the ends of the wire.  In this case it's good to look at the table for rough guidance but also calculate the IR drop.  You might end up using smaller wire than the table suggests because you can get away with it, but low voltages you might want thicker wire or bus bars to keep the voltage drop OK.

[jonpaul]

bonjour à tous

So, I just saw this thread now. 

The wire charts mentioned,  show the regulations (eg NEC, IEC) for electrical wiring eg 240V in a cable or conduit. The ratings dépend on ambient temp, environmental ( open air, under water, submersible in oil) and differ by region.  Various short term overloads are OK, eg 150% rated for 30 seconds.

So,  a consumer mains cable 2m ampacity #16 GA AWG will be different than a transformer winding of closer space magnet wire, and again different for an overhead utility wire.

Fusing capacity is much higher than rated ampacity.

In long cables, or transformers, voltage drop is considered, even if wires are well within rated ampacity.
Transformers losses are often equally iron losses ( hysterysis of core) and copper losses, ohmic heat dissipation due to winding resistance.

Finally a fine guage wire can explode (EBW, Exploding Bridge Wire) if a charged high voltage capacitor is applied. see "exploding wires" Scientific American may 1 1962.   Typically a 30..42 AWG  nichrome or copper wire a few cm long is exploded by crowbar the cap with a triggered spark gap, or Krytron tube,  35 kV,  10 J energy.   Practical EBW applications include shock wave study, and precision explosive detonators, as in nuclear devices.

bon courage,

Jon

PS    As a teenager, reading Scientific American article of 1962, I built an exploding wire experiment with DIY P11 CRT scope and 20 kV HV spark coil and 1B3 valve rectification, charging a box of HV oil caps, and using a porcelain knife switch to crowbar.  Won a high school science fair first prize....

[Kokoriantz]

If the chassis max current has something to do with cooling than it should be proportional to the wire surface which is direct function of the diameter. The wire No 31 has a diameter of 0.2mm and max chassis current about 0.5A then 1mm wire should be 5 times higher 2.5A, it is 14A. This shows that heat has nothing to do in this story.

[mag_therm]

No, Resistance goes by the square of diameter at low frequency..(R is inversely proportional to D_squared)
You can read about relationship of resistance to resistivity at low frequency, then read about Joule heating.

[ejeffrey]

If the chassis max current has something to do with cooling than it should be proportional to the wire surface which is direct function of the diameter. The wire No 31 has a diameter of 0.2mm and max chassis current about 0.5A then 1mm wire should be 5 times higher 2.5A, it is 14A. This shows that heat has nothing to do in this story.

If you ask a question because you don't understand something it's best to not then tell the people answering you that they are wrong.  It's fine to ask for clarification or explain why an answer doesn't make sense to you, but if you think you already know everything then there is no need to ask questions.

The limit is definitely thermal.  You can literally read the explanation at the top of the table here:   https://www.powerstream.com/Wire_Size.htm it does not follow the simple rule you expect because you have to account for the reduction in resistance and because air convection is a nonlinear process.  These are empirical rules of thumb and not intended as a substitute for proper engineering.  While the rules are generally quite conservative you can absolutely run into extreme situations where you need larger wire than suggested by these tables.

[Kokoriantz]

"The limit is definitely thermal"
The limit is definitely magnetic field, independent the length hence the resistance of the wire and I am asking why?
There is no any mention of length or maximum temperature the wire can handle in the table.

[TimFox]

"The limit is definitely thermal"
The limit is definitely magnetic field, independent the length hence the resistance of the wire and I ask why.

Please give an external reference or your reasoning for the statement in your last sentence.
Superconductors have a "critical magnetic field" value, but I am not aware of any such parameter for normal conductors such as metallic copper.
https://en.wikipedia.org/wiki/Critical_field

[Kokoriantz]

As I mentioned, those who design high current components, fuses, capacitors, transistors etc, know about it. They are not foolish to use steel instead of gold to link the leads to the die in power transistors or chips. I couldn't find anything on line about this, this why I ask in this forum.     

[IanB]

"The limit is definitely thermal"
The limit is definitely magnetic field, independent the length hence the resistance of the wire and I am asking why?
There is no any mention of length or maximum temperature the wire can handle in the table.

Length has nothing to do with heat generation in a wire. So thermal considerations are naturally independent of length, unless you consider a very short wire with heat conduction at its ends.

[Kokoriantz]

 The resistance is inverse the square of the thickness, so for same amount of calories, the current should be 25 times, as it is cooled by 5 times more surface, then, a wire of 1mm should carry 125 times the current of 0.2mm. If so, the limit given on the table of 0.5A for 0.2mm should be 62.5A for 1mm. It is shows 14A. 

[IanB]

The resistance is inverse the square of the thickness, so for same amount of calories, the current should be 25 times, as it is cooled by 5 times more surface, then, a wire of 1mm should carry 125 times the current of 0.2mm. If so, the limit given on the table of 0.5A for 0.2mm should be 62.5A for 1mm. It is shows 14A.

Your numbers don't add up. A 1 mm diameter wire has 25x the cross-section of a 0.2 mm wire, so it could carry (in theory) 25x the current. Twenty five times 0.5 A is 12.5 A, which is close to 14 A. But since the surface area to volume ratio is less for the thicker wire, it has to be de-rated a bit, which may explain why the actual value is a bit less than 14 A.