DC current - does it travel inside the wire or just on the outside? 1

[hrc]

I know for AC, the push-pull forces the electrons to the outside of the wire (skin effect) but for DC, in a solid wire, does it only travel on the outside of the wire, or will it 'fill' the wire. The background of the question is, a busbar attached to a PCB, where a stud ties it to the "pad". From talking to some in the bus bar industry who alluded to the fact when a stud ties into a bar, the current only really carries on the outside of the stud, and the actual transfer of current happens primarily on the surface to surface contacts of the pad and the bar.

We are talking 60A @48VDC for this application.

Comments?

 

[itsmoked]

I would say that in your applicaton "DC" would by all means "fill" the conductors.

 

[IRstuff]

You need to do a more thorough analysis of your situation.

While the DC SHOULD fill a conductor, the spreading resistance of going around corners and through interfaces, etc., WILL cause non-uniform distribution of current.

TTFN

 

[Skogsgurra]

And remember that the skin effect is quite small for power frequencies - at least as long as the conductors are below a few hundred square millimetres. For DC, it is all about contact resistance as IR says.

Gunnar Englund

http://www.gke.org

 

[VEBill]

It can be illuminating to calculate the average speed of the free electrons in a wire for a given current. Calculate the density of the free electrons, calculate the coulombs per second passing a certain point, and the resultant average speed is amazingly low (usually walking speeds).

But the light still comes on when you flick the switch.

The purpose of this thought experiment is that it may change the way you visualize current flow in conductors. It will reinforce the 'voltage = pressure' analogy and highlight the importance of the empty valence positions (versus the electrons themselves).

The point for THIS thread is that the current will not be sufficient to be accurately described as 'filling' the wire.

 

[rmw]

VE1BLL,

Excellent point. What, then would be an indicator that one would have sufficient current to "fill" so to speak, the conductor? Red hot wire? Excessive voltage (pressure) drop?

rmw

 

[hrc]

The reason I was curious on this in the first place is I have a two layer PCB, with a large copper pour on the top side where the bus bar is sitting. Two #10 studs are used to compress the bus bar to the pad to ensure a good surface to surface contact. The copper pour comes from a connector that is driving out 3000W @48VDC. I did not do any pour on the bottom side of the board to tie into the studs since I was lead to believe that there would be minimal current flow thru the holes and stud to the upper layer. Was I told incorrectly?

 

[VEBill]

"What, then would be an indicator...?"

If a conductor was carrying sufficient current that 'most' of the free electrons were on average 'moving rapidly', then the conductor would be vaporized in an instant.

 

[rmw]

I've seen that. Wasn't on purpose. Thanks for the input.

rmw

 

[itsmoked]

hrc; If this thing is built look at it running with a temperature gun. If you are getting more than ummm... 10C
You might be looking at a problem down the road.

If it's not built I would not hesitate to use both sides of the board as you are entirely discounting the bolt and the back of the bussbar. Look at any HotTub controller they all do the same thing (bolt flat bar to circuit board). They use both sides. At the very least you want to use both sides to help distribute *any* heat better.

 

[Skogsgurra]

I would also use both sides and then some kind of spring washer to avoid contact pressure changes when temperature cycles up/down.

Best way of doing this is to solder a heavy duty connector or spade terminal(s) to the board. Bolt carrying around 60 A does not sound like a good idea to me. Seen much less current burn boards around bolts that have come loose after a couple of years.

Gunnar Englund

http://www.gke.org

 

[IRstuff]

Coulomb repulsion dictates that straight runs longer than, say, 5 diameters of the wire will behave like "filled" conductors.

TTFN

 

[hrc]

I managed to change the board layout to accomidate a copper pour on both sides. Heat effect is an issue as well, since on this board, tied to the bus bar are 6 3kW supplies. 18kW total across them all.

Also...good idea on the compression springs. Hadn't thought about that before, so thanks

 

[itsmoked]

Hi hrc,

May I ask what this board is running?

If I can't that's okay.

I just like to learn.

 

[VEBill]

"...will behave like 'filled' conductors."

I understand your point that the charge carriers are spread throughout the conductor volume, but it's just that the word 'filled' might leave a false impression. The only conductors that could be fairly described as being 'filled' might be some sort of superconducting nano-fibre carrying an extremely large current.

A phrase something like 'evenly distributed' is perhaps a more accurate than 'filled'.

I admit that it's a delicate distinction and pedantic point. Sorry about that...

 

[IRstuff]

That's the purpose of the quotation marks.

TTFN

 

[VEBill]

'okay'

 

[hrc]

Sorry on the delay in getting back to this post. The application is a 48VDC distribution board (current sharing power supplies all plugged into the same bus)