X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient: geda-user AT delorie DOT com
X-TCPREMOTEIP: 207.224.51.38
X-Authenticated-UID: jpd AT noqsi DOT com
Content-Type: text/plain; charset=windows-1252
Mime-Version: 1.0 (Mac OS X Mail 7.3 \(1878.6\))
Subject: Re: [geda-user] Do you put ground planes under inductors?
From: John Doty <jpd AT noqsi DOT com>
In-Reply-To: <alpine.BSF.2.00.1407221009330.45862@earl-grey.cloud9.net>
Date: Wed, 23 Jul 2014 06:51:16 -0600
Message-Id: <66F135AF-4532-4269-A709-23BDB718DEB8@noqsi.com>
References: <CAOuGh89=mQ01sbs+Ugc3sEPUNWGexEpR_VJANTg2W_3DStUzQQ AT mail DOT gmail DOT com> <alpine DOT BSF DOT 2 DOT 00 DOT 1407221009330 DOT 45862 AT earl-grey DOT cloud9 DOT net>
To: geda-user AT delorie DOT com
X-Mailer: Apple Mail (2.1878.6)
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id s6NCpOqm011802
Reply-To: geda-user AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: geda-user AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Jul 22, 2014, at 8:16 AM, Stuart Brorson <sdb AT cloud9 DOT net> wrote:

> 
> I'd go with the ground plane since it acts as a shield, and makes
> layout easy.  It's always important to have a return path for
> currents running around the board, and putting holes/slots in a ground
> plane can interrupt return paths in unpredictable ways.
> 
> However, I would make it a point to clear other components away from
> the inductor on both sides of the board.  As you say, if the
> time-varying magnetic flux induces currents in the ground plane, any
> component on the opposite side of the board from the inductor might
> feel the effects of the induced current.  In particular, a ground
> plane with non-zero impedance will experience potential drops
> underneath the inductor, so you don't want components living there.

Generally good advice. There are a number of special cases.

For shielded inductors, including self-shielded geometries like toroids and pot cores, the field leakage is small. Treat these like any other component: generally you’ll want ground plane underneath them.

For unshielded coils, it depends on the orientation and purpose of the coil. It’s good to start from the point of view of an infinite perfectly conducting ground plane without holes. The field cannot penetrate such a plane. In this case, we may use the method of images. 

If the coil axis is perpendicular to the plane, the image coil has its magnetic axis in the opposite direction, canceling the coil’s dipole moment. Since a quadrupole field falls off faster with distance than a dipole field, this helps confine the field to the vicinity of the inductor. The cost is a strong eddy current in the plane beneath the coil, reducing the coil’s inductance and Q.

If the coil axis is parallel to the plane, the image coil has its magnetic axis in the same direction, reinforcing the coil’s dipole moment. This can increase the field away from the coil by as much as a factor of two. Because the plane intercepts less magnetic flux, the eddy currents are smaller and the reduction in inductance and Q is less than for the perpendicular case.

Removing ground and power planes below the coil reduces these effects. It also allows some of the field to leak to the far side of the board. Remember that routing signal or power traces across any sort of gap in your ground plane is usually a bad idea.


John Doty              Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com