[SI-LIST] Re: Inductance and Insertion Loss
- From: andrew.c.byers@xxxxxxxxxxxxxx
- To: mmoeller@xxxxxxxxxxxxxxx, si-list@xxxxxxxxxxxxx
- Date: Thu, 15 Apr 2004 13:46:28 -0700
Merrick,
Attenuation can be defined as "loss", which is energy that does not pass to
either port 1 or 2 in a (closed) two-port system. This energy can be
tranformed to heat in the metal carrier (conductor loss) or heat in the
dielectric (dielectric loss). In an assumed two-port system, you might also
have loss due to radiation (consider this another port), crosstalk coupling
(to another port), or conversion to another mode. For example, we have seen
many posts about how unmatched differential lines will convert
differential-mode signals to a certain percentage of common-mode components.
If you were looking at the S-parameters of the differential mode, you would
find that not all the energy you put into port 1 can be accounted for in
that mode alone.
That said, your case of inductance causing insertion loss (S21) makes sense
when one of these attenuation-causing events happens (which probably does).
A good way to check if you have bonafide LOSS as opposed to reflection (S11)
is to check this simple equation:
[mag(s11)]^2 + [mag(s21)]2 = 1.
If this holds, then you have a lossless, mode-conversionless system, and all
your S21 variations are due to the signal reflecting off the impedance
interface between your chip tline and the connector pin. This is easy to
model and check.
If the value above is less than one, then you have crosstalk (physical or
modal, including radiation) and/or loss (conductor or dielectric). The
actual inductance of the pin might affect the crosstalk components, but not
the loss components (unless more inductance means a thinner pin, which will
have more resistance and more loss, but the length of the pin probably has a
bigger impact on inductance...). Basically, your crosstalk can be controlled
by physical design and layout, and the loss can be controlled by material
choice. The terms "wave coupling" or "energy coupling" probably refer to a
modal conversion that happens. In your case (especially if it is
single-ended) I would expect some kind of cavity-mode or parallel plate mode
to be triggered by a transition between environments (imagine a radiating
wave effect coming off your pin, much like an antenna...). The inductance
will definitely impact the amount of energy that gets coupled into this
other mode. Shortening up the pin will drive that coupling point higher up
in frequency, to be worried about in the future.
A good way to see this effect would be to run an EM solver (HFSS, or FDTD)
and watch the energy travel from the chip down the pin to the package. When
it hits the pin you should see a small amount of energy spread out and
couple into other pins. I am sure the EM solver vendors have examples that
illustrate this - I used FDTD to study this effect back in grad school (ah,
when we had time to follow our whims and discover things).
Cheers,
Andrew Byers
-----Original Message-----
From: Moeller, Merrick [mailto:mmoeller@xxxxxxxxxxxxxxx]
Sent: Thursday, April 15, 2004 10:07 AM
To: Si-List (E-mail)
Subject: [SI-LIST] Inductance and Insertion Loss
Experts,
It has been brought to my attention that the inductance of connector
pins in a LGA type chip to board contact, is considered more for reasons of
insertion
loss than any other reason. It has been explained to me that given a system
with
a loss less dielectric housing the attenuation seen on an active line will
be due to,
what may be called "energy coupling" or "wave coupling". In understanding
the
principle of crosstalk on coupled lines, and the relationship of frequency
to the
amount of coupled voltage it seems that insertion loss of a contact in an
ideal environment
would be explained by the amount of crosstalk.
Can insertion loss in an ideal system be contributed to simple
crosstalk, or
do the wave effects determine the loss on a more complicated scale? If so,
how does the
inductance of the contacts effect the wave effects?
I hope this isn't too vague. I have looked quite a bit and found no
direct correlation
between inductance and insertion loss. Can you help make the connection?
Regards,
Merrick
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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
------------------------------------------------------------------
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 FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
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:
