[SI-LIST] Gigabit fiber transceivers, EMI, trace lengths, and... antennas.

  • From: "Kolstad, Joel (EIP)" <jkolstad@xxxxxxxxxxxxxxx>
  • To: "'si-list@xxxxxxxxxxxxx'" <si-list@xxxxxxxxxxxxx>
  • Date: Thu, 9 Aug 2001 14:19:58 -0700 

Hi,

We have a board that uses an Amp 269152-1 gigabit fiber transceiver; we use
it at 1.0625Gbps (Fibre Channel speeds) with an industry-standard SerDes
operating at 106.25MHz.  This is mounted on a standard PCI card, with the
nose of the transceiver poking through the rear bracket of the PCI card.
What we've found during CE compliance testing is that the little metal
"clip" around the transceiver's nose needs to be in close physical contact
to the rear bracket, or else we significantly exceed emissions standards...
at 318.75MHz.  So, we have a fix, but I'm curious -- how does a 318.75MHz
signal radiate from a hole in the rear bracket that's ~.4"x.3" wide?  (Since
the wavelength of a 318.75MHz signal is in the ballpark of a meter I was
under the impression that, if you think of the opening in the bracket as a
waveguide, that you're nowhere near any supported mode of the waveguide.)

Also... in the past couple of days, it was stated that for 2.5Gbps
transceivers, striplines are recommended as opposed to microstrip due to
their lower radiation.  Fair enough.  Some of the commentary, however, also
seemed to suggest that a quarter-wave microstrip transmission line would
make an excellent antenna.  Is that really true?  It seems to me that if
you're building a standard microstrip line (copper trace above a ground
plane), sure, it'll certainly radiate some (and more so than stripline
construction, even), but to make a _really_ good antenna you need to remove
the ground plane so that (most of) the EM field isn't just immediately
cancelled by the fields due to the return current on the ground plane.

OK, ok, one final question -- we know that a standard dipole has an
increasing radiation resistance up to its resonant frequency, at which point
the length of each half of the dipole is approximately a quarter wavelength.
If we make the dipole antenna longer, the radiation resistance will actually
decrease (right?), and in general the plot of the radiation resistance vs.
antenna length is a cyclical function.  Is it correct to think of the
decreasing radiation resistance as being due to partial field cancellation
once the antenna is longer than its resonant length?  I.e., if each arm of
the dipole is a half wavelength long, it's correct to think of this as
really two quarter-wave dipoles stacked on top of each other: The fields
from the "upper" quarter-wave dipole end up being 180 degrees out of phase
with those from the "lower" quarter-wave dipole, largely canceling each
other out, and therefore there's little radiation whatsoever with such an
"antenna."

(I'd be curious if anyone has a suggestion for a book that stresses the
intuitive understanding of a lot of this subject matter.  I did OK in my
college fields and waves courses, but I find that I'm rather lacking in
intuition as far as how, e.g., a folded dipole or quadrifilar helix antenna
ought to "behave" [radiate] vs. a simple dipole.  Even though these days you
can certainly stick whatever arbitrary configuration you want into a program
like NEC and "see what happens," I'd bet a nickel that the good antenna
designers out there can tell you off the top of their heads what an
antenna's going to do -- to a first order approximation -- before they ever
touch a PC.  Is that true?)

Oh, wait, this is the SI-List, not Antenna List... sorry for getting a
little off-topic. :-)

---Joel Kolstad
Electroglas
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
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:

  1. Home
  2. si-list
  3. Archive
  4. 08-2001