[SI-LIST] Re: si-list Digest V3 #194
- From: "Sainath Nimmagadda" <gigabit@xxxxxxxxxx>
- To: <andrew.c.byers@xxxxxxxxxxxxxx>
- Date: Fri, 18 Jul 2003 17:06:32 -0800
Andy,
Yes, the inductance value should remain the same for both cases. Also,
we are capturing all the magnetic flux lines in both cases.
Now comes the real question. When you capture all the flux lines, is the
inductance going to be maximum? or minimum?
Sainath
---------Included Message----------
>Date: Fri, 18 Jul 2003 09:50:57 -0700
>From: <andrew.c.byers@xxxxxxxxxxxxxx>
>Reply-To: <andrew.c.byers@xxxxxxxxxxxxxx>
>To: <gigabit@xxxxxxxxxx>
>Subject: RE: [SI-LIST] Re: si-list Digest V3 #194
>
>Sainath -
>
>With the case of the surface above the microstrip, the inductance
value
>should remain the same. The integrating distance will be from the top
of the
>microstrip to infinity, and the B-field will be diminishing in
magnitude as
>you get further and further from the microstrip. The integral to
infinity
>will be equivalent to some series, and can be solved easily to a
finite
>number.
>
>Another way of looking at it - all of the fields that wrap under the
>microstrip will also wrap above it. You just have to have a big enough
>surface to catch them all. In practice, a surface that is about 3-4
times
>the height of the dielectric should catch most of the fields. This
whole
>infinite surface stuff is just for theoretical robustness.
>
>By the way, there is a paper that demonstrates this in FDTD simulation.
I
>believe it is in the 1997 EPEP conference - its written by Melinda
Piket-May
>and Roger Gravrok. I might be off by a year of two... if you have
those
>conference proceedings look for it. I can dig more for the name if you
would
>like.
>
>andy
>
>-----Original Message-----
>From: Sainath Nimmagadda [mailto:gigabit@xxxxxxxxxx]
>Sent: Thursday, July 17, 2003 11:44 PM
>To: Byers, Andrew C
>Cc: si-list@xxxxxxxxxxxxx
>Subject: RE: [SI-LIST] Re: si-list Digest V3 #194
>
>
>Hi Andy,
>
>Thanks again. I get the themes that inductance is a one number affair
>and current returns through the least inductance path. Is there a
>contradiction in these themes?
>
>Let me borrow the following from your previous mail.
>
>"If you were to put your integrating surface on the other side of the
>trace, extending up from the top of the trace, you theoretically would
>have to make the area of the surface extend to infinity to "catch" all
>the field lines."
>
>For this case, is the inductance of the microstrip going to be
>infinity(because of infinite surface)? or any other value? remains same
>as what it was for the integrating surface below the trace?
>
>Sainath
>
>
>
>
>---------Included Message----------
>>Date: Thu, 17 Jul 2003 17:37:12 -0700
>>From: <andrew.c.byers@xxxxxxxxxxxxxx>
>>Reply-To: <andrew.c.byers@xxxxxxxxxxxxxx>
>>To: <gigabit@xxxxxxxxxx>
>>Cc: <si-list@xxxxxxxxxxxxx>
>>Subject: RE: [SI-LIST] Re: si-list Digest V3 #194
>>
>>Hello Sainath,
>>
>>Clearing up some terminology here.
>>
>>"Least inductance" refers to the path that the current will travel
>because
>>it has the least inductance of all possible paths in the system.
>Current
>>will never choose an alternate path of "most inductance". BUT you can
>have a
>>different design in which the "path of least inductance" is longer.
>For
>>example a two wire line with no ground plane where the wires are
>extremely
>>far apart. Huge loop, huge inductance. But still the smallest loop for
>that
>>system. For a microstrip, a path of More Inductance would be if there
>were a
>>gap in the ground plane under the microstrip line. The current would
>be
>>forced to diverge around the gap. This path would be more inductive
>than a
>>solid ground plane, but the current would still be following the path
>of
>>least inductance for that particular case.
>>
>>The main challenge in most systems I've dealt with is making sure
that
>>return current paths have the least inductance possible. The simplest
>way to
>>do this is go differential. Then you carry your virtual ground with
>you
>>everywhere. If single ended, then be very conscious about where the
>return
>>currents flow and try to provide a short path. Plenty of threads on
>this
>>list about that.
>>
>>Not sure if this clears up your last question, hope it helps though.
>>
>>- Andy
>>
>>
>>
>>-----Original Message-----
>>From: Sainath Nimmagadda [mailto:gigabit@xxxxxxxxxx]
>>Sent: Thursday, July 17, 2003 4:01 PM
>>To: Byers, Andrew C
>>Cc: si-list@xxxxxxxxxxxxx
>>Subject: RE: [SI-LIST] Re: si-list Digest V3 #194
>>
>>
>>Andy,
>>
>>Thanks. I appreciate the extra effort to explain detail of
>integration.
>>In short, you've explained the current loop formed by a signal path on
>
>>trace and signal return path beneath the trace and on the ground
plane.
>
>>Such a return path, with its minimum loop area, is widely known to
>>provide the path of "least" inductance for high-frequency currents(for
>
>>example, Black Magic book). If inductance is thought of as one number,
>
>>what does "least inductance" refer to? Which is the path of "most"
>>inductance for the microstrip? No doubt, I'm missing somethig.
>>
>>Sainath
>>
>>---------Included Message----------
>>>Date: Thu, 17 Jul 2003 10:02:49 -0700
>>>From: <andrew.c.byers@xxxxxxxxxxxxxx>
>>>Reply-To: <andrew.c.byers@xxxxxxxxxxxxxx>
>>>To: <gigabit@xxxxxxxxxx>, <beneken@xxxxxxxxxxxx>
>>>Cc: <si-list@xxxxxxxxxxxxx>
>>>Subject: RE: [SI-LIST] Re: si-list Digest V3 #194
>>>
>>>Sainath,
>>>
>>>As Thomas pointed out, inductance is the ratio of magnetic flux to
>>current
>>>in the conductor. Magnetic flux is the integral of B dot dA, or the
>>magnetic
>>>field [dot product] the surface you are integrating over. The "dot
>>product"
>>>is the same as multiplying the B-field by the area by the cosine of
>>the
>>>angle between the B-vector and the normal to the area. So if the
>>B-vector is
>>>perpendicular to the area surface, then the B-vector is parallel to
>the
>>unit
>>>normal vector of the area surface, cosine of this zero degree angle
is
>
>>1,
>>>and you simply multiply B*area. Here's an example to illustrate.
>>>
>>>You have a rectangular metal trace over a ground plane, length in
the
>>>z-direction, height in the y, width in the x. Stretch a rectangle in
>>the yz
>>>plane between the trace and the ground plane. Make it any length
>>(smaller if
>>>you are simulating with EM tool). If we assume perfect conductors (ie
>
>>no
>>>internal-conductor magnetic fields), then all of the magnetic field
>>>associated with that signal trace will pass through this rectangle.
It
>
>>is
>>>kind of like a net. Magnetic field lines always have to end up in the
>
>>same
>>>place they started, completing the circle. Also, in this
>configuration,
>>all
>>>your field lines are perpendicular to the integrating rectangle. So
>>>inductance is flux/I = B*A/I. In this case, you will actually have
>>>inductance per unit length because your net had a specific z-length.
>>>
>>>If you were to put your integrating surface on the other side of the
>>trace,
>>>extending up from the top of the trace, you theoretically would have
>to
>>make
>>>the area of the surface extend to infinity to "catch" all the field
>>lines.
>>>By placing it between the signal line and the return path, you
capture
>
>>all
>>>the field lines. So you have one number for inductance if you account
>
>>for
>>>all the B field lines. An inductance "distribution" would indicate
>that
>>you
>>>are not catching all the magnetic field lines with your integrating
>>surface.
>>>
>>>
>>>This might open up a talk about internal inductance, when you have
>>magnetic
>>>field lines (ie current) INSIDE the conductors. As frequency
>increases,
>>the
>>>current crowds to the surface, and the internal inductance
diminishes.
>
>>But
>>>at lower or intermediate frequencies, this internal inductance can be
>
>>a
>>>contributing factor. For PCB's, this is typically in the low MHz
>range.
>>But
>>>for square conductors on silicon, measuring a few microns wide and a
>>few
>>>microns high, the internal inductance might have to be considered up
>>to
>>>several GHz. Does this affect you? Do you electrical models consider
>>this
>>>effect? How about internal inductance of the ground plane?
Interesting
>
>>stuff
>>>here.
>>>
>>>Salud,
>>>
>>>Andy Byers
>>>
>>>-----Original Message-----
>>>From: Sainath Nimmagadda [mailto:gigabit@xxxxxxxxxx]
>>>Sent: Thursday, July 17, 2003 9:25 AM
>>>To: beneken@xxxxxxxxxxxx
>>>Cc: si-list@xxxxxxxxxxxxx; gigabit@xxxxxxxxxx
>>>Subject: [SI-LIST] Re: si-list Digest V3 #194
>>>
>>>
>>>Thomas,
>>>
>>>Thank you. I agree, you get one value of inductance for one
>>integration.
>>>If you repeat this for a number of 'concentric spheres', you will get
>a
>>
>>>number of inductances- ranging from minimum to maximum. Does that
make
>
>>
>>>sense?
>>>
>>>Sainath
>>>
>>>---------Included Message----------
>>>>Date: Thu, 17 Jul 2003 12:04:57 +0200
>>>>From: "Thomas Beneken" <beneken@xxxxxxxxxxxx>
>>>>Reply-To: <beneken@xxxxxxxxxxxx>
>>>>To: <si-list@xxxxxxxxxxxxx>
>>>>Subject: [SI-LIST] Re: si-list Digest V3 #194
>>>>
>>>>Hello Sainath,
>>>>
>>>>inductance is the proportional factor between the current and the
>>>magnetic
>>>>flux. So far Your idea is ok. But calculating magnetic flux from
>>>magnetic
>>>>field requires an integration across a closed surface surrounding
>the
>>>>conductor carrying the current. So - as You see - You will not get
a
>>>>inductance distribution over conductor length but only an integral
>>>value for
>>>>the conductor enclosed in the chosen sphere.
>>>>
>>>>Sorry,
>>>>Thomas
>>>>
>>>>> Msg: #12 in digest
>>>>> Date: Wed, 16 Jul 2003 11:55:35 -0800
>>>>> From: "Sainath Nimmagadda" <gigabit@xxxxxxxxxx>
>>>>> Subject: [SI-LIST] Microstrip Inductance
>>>>>
>>>>> Hello experts:
>>>>>
>>>>> For a microstrip, we know the magnetic field distribution(for
>>>>> example,
>>>>> Fig. 2.3 Stephen Hall's book) and current density
>>>>> distribution(Fig. 4.5
>>>>> same book). Given these, how would you obtain the inductance
>>>>> distribution?
>>>>>
>>>>> Thanks in advance,
>>>>> Sainath
>>>>
>>>>
>>>>------------------------------------------------------------------
>>>>To unsubscribe from si-list:
>>>>si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject
>field
>>>>
>>>>or to administer your membership from a web page, go to:
>>>>//www.freelists.org/webpage/si-list
>>>>
>>>>For help:
>>>>si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>>>>
>>>>List archives are viewable at:
>>>> //www.freelists.org/archives/si-list
>>>>or at our remote archives:
>>>> http://groups.yahoo.com/group/si-list/messages
>>>>Old (prior to June 6, 2001) list archives are viewable at:
>>>> http://www.qsl.net/wb6tpu
>>>>
>>>>
>>>>
>>>---------End of Included Message----------
>>>_____________________________________________________________
>>>
>>>
>>>------------------------------------------------------------------
>>>To unsubscribe from si-list:
>>>si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject
field
>>>
>>>or to administer your membership from a web page, go to:
>>>//www.freelists.org/webpage/si-list
>>>
>>>For help:
>>>si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>>>
>>>List archives are viewable at:
>>> //www.freelists.org/archives/si-list
>>>or at our remote archives:
>>> http://groups.yahoo.com/group/si-list/messages
>>>Old (prior to June 6, 2001) list archives are viewable at:
>>> http://www.qsl.net/wb6tpu
>>>
>>>
>>---------End of Included Message----------
>>_____________________________________________________________
>>
>>
>---------End of Included Message----------
>_____________________________________________________________
>
>
---------End of Included Message----------
_____________________________________________________________
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
