[SI-LIST] Re: Mode conversion

Ben,

Your first question is an interesting one.  If you have the situation where
you have two differential transmission lines with the same differential mode
impedances, but with different common-mode impedances, then you ask if when
they are cascaded, does this mismatch in common-mode impedances create the
conversion of differential signals to common-mode signals?  This is a
hypothetical and somewhat idealized situation.  The answer is no, a
common-mode impedance mismatch does not create imbalance, hence it does not
cause mode conversion.  But, you will find that the common-mode signal that
exists on the first line (and it always exists in real life) will see an
impedance mismatch at the junction of the two transmission lines and will
undergo a reflection.  The reflection may increase the common-mode noise
characteristics of your system of transmission lines, but is unlikely to
effect your differential signal.  In other words, the common-mode reflection
may be more of an EMI issue than a signal integrity issue.

One beauty of the differential transmission line is that the differential
signal is very loosely coupled to the common-mode signal.  What I mean by
that is that you can introduce significant imbalance and produce a lot of
common-mode, but you will find that the differential signal is only slightly
effected.  I can show you some evidence of this, if you are interested.

Mode Conversion in a connector:  As Julian Ferry pointed out in another
posting, there are "environmental" features in a connector that can be
different in one side of the line vs. the other and cause the imbalance that
results in mode conversion.  A difference in path length through the
connector is only one of several ways that this can happen.

Along this line, you should realize that two parallel traces of equal length
on a PCB built from common FR-4 material can experience imbalance.  Fiber
glass board material is inhomogeneous because the glass weave is not the
same all over the board.  So, you find that the electrical properties of the
board can be a little different in the vicinity of one trace vs. the other.
This gives you an inherent mode conversion mechanism in a real board that
would appear to be very well balanced in terms of trace lengths. This
phenomenon becomes more important as frequency increases.

Hope this is helpful,

Jim



Jim Knighten, Ph.D.
Teradata, a Division of NCR             http://www.ncr.com
17095 Via Del Campo
San Diego, CA 92127
USA
Tel: 858-485-2537
Fax: 858-485-3788
jim.knighten@xxxxxxx

 -----Original Message-----
From:   Ben Rothchild [mailto:benrothchild@xxxxxxxxx] 
Sent:   Thursday, February 21, 2002 1:31 AM
To:     si-list@xxxxxxxxxxxxx
Subject:        [SI-LIST] Re: Mode conversion


Jim, Doug, Ken, Mick,

Thanks for your explanation on Mode Conversion.

And to all Si Experts.

Is there a possibility of mode conversion when
a differential signal passes from one set of 
ideal(equal length etc) differential lines with 
Zodd-1 and Zeven-1 impedances to another set
of ideal differential lines with Zodd-2 and Zeven-2
impedances where Zodd-1 = Zodd-2.

Second is there a possibility of mode conversion
when passing through a connector other than due to
differences of pin lengths. Ofcourse coupling
in the connector could be different than
the transmisssion line and its effect on mode
conversion will I suppose be the same explanation 
as for the case of two sets of differential lines
mentioned above.

Any insights/explanations

Ben

> A simple explanation that I've always liked is a
> diff line with unequal
> "lengths" (physical and/or mode induced)causes the
> currents to have
> different phases or arrival times, and therefore the
> currents do not cancel
> completely, i.e., are not exactly 180 phase from
> each other.  Due to the
> high impedance of the receiver a very small amount
> of current can develop a
> large common mode voltage across the receiver input,
> producing common mode
> radiation.  Microstrips have more issues than
> striplines due to
> inhomogeneous permittivties and surface effects.  I
> haven't found right
> angle "corner" capacitance to have any measureable
> effect ....yet.  If
> meander sections are tightly coupled, you could have
> issues even at slow
> speeds.  If you lay them out correctly they pose no
> problem.
> 
> -----Original Message-----
> From: si-list-bounce@xxxxxxxxxxxxx
> [mailto:si-list-bounce@xxxxxxxxxxxxx]On Behalf Of
> Doug McKean
> Sent: Tuesday, February 19, 2002 4:20 PM
> To: Signal Integrity Group
> Subject: [SI-LIST] Re: Mode conversion
> 
> 
> 
> In terms of EMI, the full name for "mode conversion"
> is
> "differential-to-common mode noise conversion". 
> It's
> when differential mode noise gets converted to
> common
> mode noise. The exact mechanics of this escapes me,
> but does happen to varying degrees with just about
> any differential buss layout.  The worse the layout,
> the
> more common mode noise produced. For instance,
> put lots of equal to or greater than 90 degree bends
> in
> a differential buss layout between points A and B
> and
> you will get lots more common mode noise as opposed
> to
> a straight buss configuration between the same two
> points.
> Not sure if it's some odd/even mode phenonmenon or
> some additive crosstalk thing. I'm sure someone here
> would know.  - Doug McKean
> 
> 
>
------------------------------------------------------------------
> 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:
> http://www.freelists.org/webpage/si-list
> 
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the
> Subject field
> 
> List archives are viewable at:
>               http://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:
> http://www.freelists.org/webpage/si-list
> 
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the
> Subject field
> 
> List archives are viewable at:     
>               http://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
>   
> 


__________________________________________________
Do You Yahoo!?
Yahoo! Sports - Coverage of the 2002 Olympic Games
http://sports.yahoo.com
------------------------------------------------------------------
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:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                http://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:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                http://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: