Ken,
May be I don't understand what you are saying. My
apologies. I thought you were saying that a floating
shield is only effective against a magnetic field source
*if* that floating shield is larger than the wavelenth of
the frequency it is trying to shield, which of course is
not true.
I thought you were saying that if, say, I have a loop of
wire that is 1/2 inch in diameter carrying 200mA at 100MHz
[which of course will cause you to fail radiated Class B]
that when I put a 2 inch plate of floating conductive
material over this loop that it will not effectively do any
shielding, because I *must* shield with much larger than
the wavelength of 100 MHz, which is around 118 inches, or
10 feet?
- Robert -
On Thu, 15 Sep 2005 19:33:09 -0500
Ken Javor <[email protected]> wrote:
> There is not much point to further discussion, because I
> just keep repeating
> myself. The case you are discussing is partially valid
> where the shield
> material is large relative to a wavelength, and it is
> reflecting an
> electromagnetic field. The currents induced in this case
> are due to the
> electric field vector, according to J = sigma E. But
> even here, without an
> adequate shield termination the currents induced to flow
> on the inside of
> the shield will escape to the outside and re-radiate.
>
> If the shielding is specifically for the low frequency
> magnetic component of
> the field then the eddy currents flow due to Faraday Law
> induction, which is
> a vector phenomenon, and then the shield has to form a
> loop penetrated by
> the magnetic field vector, or it won't work.
>
> > From: "Robert A. Macy" <[email protected]>
> > Date: Thu, 15 Sep 2005 15:27:54 -0700
> > To: Ken Javor <[email protected]>
> > Cc: [email protected]
> > Subject: Re: When is a cable shield not a cable shield?
> >
> > Eddy currents exist everywhere the conductor looks like
> a
> > shorted turn on a transformer.
> >
> > Imagine a perfectly balanced pair of wires, driven in a
> > perfetly balanced manner. A wire down, a wire back - a
> > loop. That pair will radiate right out through its
> loop.
> > A simple plate of conductor placed over the loop will
> get
> > eddy currents generated in it that "cancels" to lower
> > radiation.
> >
> > The plate does not have to be terminated to have this
> > effect. Simply covering the surface of the loop is
> > sufficient. Plus with the wires perfectly balanced and
> the
> > drive pervectly balanced there will be no residual
> signal
> > put upon the plate, so the plate does not radiate, or
> > reradiate.
> >
> > - Robert -
> >
> > On Thu, 15 Sep 2005 14:41:56 -0500
> > Ken Javor <[email protected]> wrote:
> >> See response to Mr. Woodgate. Shorted fibers are not
> an
> >> issue, my argument
> >> holds if the shield material is a solid copper pipe.
> You
> >> won't get eddy
> >> currents flowing in that pipe until the ends are
> >> terminated so currents can
> >> flow in a loop. The magnetic vector flows around the
> >> circumference of the
> >> pipe, but that is not the path the eddy currents nee
> to
> >> take. They must
> >> flow longitudinally down the pipe. The cable becomes
> the
> >> center of a
> >> coaxial transmission line; the shield is the coaxial
> >> transmission line
> >> shield and return conductor.
> >>
> >>> From: "Robert A. Macy" <[email protected]>
> >>> Date: Thu, 15 Sep 2005 11:28:26 -0700
> >>> To: Ken Javor <[email protected]>
> >>> Cc: [email protected]
> >>> Subject: Re: When is a cable shield not a cable
> shield?
> >>>
> >>> Yes, it can.
> >>>
> >>> The fibres are all shorted out making it a
> pseudo-metal
> >>> plane.
> >>>
> >>> Plus, the loops of interest are pretty much contained
> >> in
> >>> between the two ends of the shield.
> >>>
> >>> To see how much effect there is measure the
> inductance
> >> of
> >>> the loop inside shield and ouside the loop.
> >>>
> >>> - Robert -
> >>>
> >>> On Thu, 15 Sep 2005 11:50:33 -0500
> >>> Ken Javor <[email protected]> wrote:
> >>>> I have a problem with this as well. A non-magnetic
> >>>> shield material protects
> >>>> against magnetic field radiation by providing a path
> >> for
> >>>> eddy currents to
> >>>> flow in. The eddy currents flow in such a way as to
> >>>> cause a magnetic field
> >>>> that opposes the originating field (Lenz' law). In
> >> order
> >>>> for the eddy
> >>>> currents to have sufficient magnitude to cause field
> >>>> cancellation, the
> >>>> shield must provide a very low impedance short
> >> circuit.
> >>>> An open-circuited
> >>>> shield braid cannot perform this function.
> >>>>
> >>>>> From: "Robert A. Macy" <[email protected]>
> >>>>> Date: Thu, 15 Sep 2005 09:16:09 -0700
> >>>>> To: "Hudson, Alan" <[email protected]>
> >>>>> Cc: [email protected]
> >>>>> Subject: Re: When is a cable shield not a cable
> >> shield?
> >>>>>
> >>>>> Alan,
> >>>>>
> >>>>> Think also in terms of magnetic fields, or "area
> >>>> loops".
> >>>>>
> >>>>> The area defined by the small loop between signal
> >> lines
> >>>> and
> >>>>> ground lines now have rather opaque metal plates
> over
> >>>> them.
> >>>>> Starting around 10KHz the conductive "shield" kills
> >> all
> >>>>> the magnetic fields coming out. Which were
> probably
> >>>> more
> >>>>> significant than the E-Field contributions.
> >>>>>
> >>>>> - Robert -
> >>>>>
> >>>>> On Thu, 15 Sep 2005 16:22:31 +0100
> >>>>> "Hudson, Alan" <[email protected]> wrote:
> >>>>>> G'Day!
> >>>>>>
> >>>>>> Came across an oddity today which has failed to
> >>>> stimulate
> >>>>>> any memories (from
> >>>>>> about the time the old King died!) of basic
> >>>>>> electromagnetic theory. A
> >>>>>> well-known computer Company supplies what it calls
> >>>>>> "shielded ribbon cables".
> >>>>>> When we examine one of these we find it's
> >> essentially
> >>>> a
> >>>>>> metallic braid
> >>>>>> sleeve around a standard ribbon cable. The braid
> >> isn't
> >>>>>> connected to anything
> >>>>>> (not even a drain wire at the connector ends).
> >>>>>>
> >>>>>> So what level of shielding is this going to give -
> >> if
> >>>>>> any?
> >>>>>>
> >>>>>> I could imagine signal lines being capacitively
> >>>> coupled
> >>>>>> to the braid and
> >>>>>> then capacitively coupled to any 0V lines in the
> >>>> ribbon.
> >>>>>> But I can also
> >>>>>> imagine signal lines being capacitively coupled to
> >>>> other
> >>>>>> signal lines.
> >>>>>> Similarly with external signals hitting the braid
> -
> >>>> they
> >>>>>> could go to either
> >>>>>> the signal or 0V lines - yes/no?
> >>>>>>
> >>>>>> I know I should know the answer to this, but
> >>>> obviously,
> >>>>>> today at least, if I
> >>>>>> had a brain I'd be dangerous!
> >>>>>>
> >>>>>> :-)
> >>>>>>
> >>>>>> Regards,
> >>>>>>
> >>>>>> Alan
> >>>>>> --
> >>>>>> Principal Engineer,
> >>>>>> BAE Systems Limited,
> >>>>>> UK
> >>>>>
> >>>>> -
> >>>>> 2005 IEEE Symposium on Product Safety Engineering
> >>>>> 3-4 October Schaumburg, IL
> >>>>> http://www.ieee-pses.org/symposium
> >>>>>
> >
>
> -
> 2005 IEEE Symposium on Product Safety Engineering
> 3-4 October Schaumburg, IL
> http://www.ieee-pses.org/symposium
>
-
2005 IEEE Symposium on Product Safety Engineering
3-4 October Schaumburg, IL
http://www.ieee-pses.org/symposium
This message is from the IEEE Product Safety Engineering Society
emc-pstc discussion list. Website: http://www.ieee-pses.org/
To post a message to the list, send your e-mail to [email protected]
Instructions: http://listserv.ieee.org/listserv/request/user-guide.html
List rules: http://www.ieee-pses.org/listrules.html
For help, send mail to the list administrators:
Scott Douglas [email protected]
Mike Cantwell [email protected]
For policy questions, send mail to:
Richard Nute: [email protected]
Jim Bacher: [email protected]
All emc-pstc postings are archived and searchable on the web at:
http://www.ieeecommunities.org/emc-pstc
