[SI-LIST] Re: Diff line
- From: Scott McMorrow <scott@xxxxxxxxxxxxxxxx>
- To: benrothchild@xxxxxxxxx
- Date: Fri, 15 Feb 2002 09:29:06 -0800
Ben,
Here are the answers to your questions.
>How is impedance control dependent on whether
>they are loosely/tightly coupled.
>
With loosely coupled differential pairs, manufacturing etch contol
affects the width of the traces, which alters the impedance. For
tightly coupled differential pairs, the width between the pairs is also
greatly altered by the etch dimensions. Impedance of a differential
pair trace is controled by the width of the trace and the separation of
the edges. The percentage change for both is higher with a tightly
coupled differential pair.
>What is even mode return loss and how is coupling
>affecting this.
>
Even mode return loss is S11 for the common mode signal (the part of the
signal that is not differential and is caused by driver and delay skew)
and is the reflected energy that returns back to the driver. Higher
return loss means more common mode energy is bouncing around. This is
usually due to impedance mismatch in the system. Tighter trace coupling
causes a larger divergence in odd and even mode impedances. If the
system is matched for odd mode (differential transmission) then the even
mode suffers greater mismatch and return loss. This can become a
problem when the even mode return loss sets up unwanted resonance
conditions in the system.
>How is insertion loss affected by coupling.
>
To achieve the same differential impedances, conductor width must be
smaller for tightly coupled differential pairs, when compared to loosely
coupled differential pairs. This smaller width causes higher DC
conductor loss and higher AC losses at high frequencies due to skin
effect. In addition, a larger percentage of the fields for tightly
coupled differential pairs are concentrated towards the edge of the
pair, increasing the field concentration here and thereby increasing the
conductor losses.
>Are you experts there driving gigabit links
>through 2mm open pin field connectors!
>Its very hard to believe. Or is it controlled
>impedance connectors. But then are there any
>controlled impedance 2mm connectors. Any
>inputs on what connectors you are using and
>the highest data rate.
>
Teradyne, Molex, Tyco, FCI, Erni and many others make a number of
varieties of 2 mm connectors which have controlled impedance for 50 ohm
single-ended and 100 ohm differential signals. We've used them
successfully for 2.5 and 3.125 Gbps system backplanes. Some of the
newer connectors are capable of supporting nearly 10 Gbps data transmission.
>How is impedance variation at connector, BGA
>escapes and via transitions affected by
>coupling.
>
Tightly coupled differenential pair separation must be widened at the
point of insertion into these strucures. Changing the separation of the
pairs causes a change in impedance.
>What resonance are you refering to, could you
>please elaborate.
>
Whenever an impedance mismatch occurs a resonance structure is formed.
Quarter-wave and half-wave resonance structures are created by these
mismatches. These resonance modes can cause degradation in signal
waveshape and can couple to other similar resonance structures across
the board. I'd refer you to the book: Microwave Engineering by Pozar
for more information on resonance.
Good questions, Ben.
regards,
scott
--
Scott McMorrow
Principal Engineer
SiQual Interconnect Engineering
18735 SW Boones Ferry Road
Tualatin, OR 97062-3090
(503) 885-1231
http://www.siqual.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:
//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:
