[SI-LIST] Re: loosely vs tightly coupled differential pairs

Hello David,

Some thoughts on this.
You have to take a look at all of the parameters that drive your decision.
For example.  What are the physical constraints you're up against in your
design?  You might have pin or via fields that do not allow two traces
between and thus you would have to "uncouple" a coupled pair.

A tightly coupled line concentrates its fields to its complementary pair.
You have to consider other parameters such as trace to plane separation.
But in general this is the case with tightly coupled lines.

In loosely coupled lines you want to keep the pairs separated such that
crosstalk does not adversely affect your neighbor signals.  You will still
have crosstalk from one line to another regardless of your choice
(tight/loose).  You will need to analyze the crosstalk parameters when
choosing your physical arrangement.  Add that to your system budget and see
how much xtalk you can tolerate.

In loosely coupled routing you can allow the signals to take separate paths
but you don't want to introduce noise through coupling of surrounding
objects.  You might consider signal/plane/signal routing and keep the paths
mirrored.

I would recommend you get your hands on a layout tool to be able to
implement and test route various strategies.  And a good 2D tool if you want
to consider crosstalk.  My recommendation is Eagleware's Genesys for this.
It's very straightforward to use.  You can look at xtalk through
s-parameters.

So... don't choose blindly.  Attack the problem several ways and come upon
the natural solution.

Hope this helped.

Bill

> -----Original Message-----
> From: si-list-bounce@xxxxxxxxxxxxx
> [mailto:si-list-bounce@xxxxxxxxxxxxx]On Behalf Of David Kaiser
> Sent: Friday, May 10, 2002 3:09 PM
> To: 'si-list@xxxxxxxxxxxxx'
> Subject: [SI-LIST] loosely vs tightly coupled differential pairs
> 
> 
> 
> I want to get some opinions about edge coupled differential traces.
> 
> Which is better? Loosely coupled or tightly coupled 
> differential traces.
> How many times the trace width seperation is necessary to be 
> concidered
> loose or tight?
> 
> - I already know that loosely couple traces may be difficult to route
> through dense via fields.
> - I also have been told that tight coupling lowers EMI.
> - A tightly coupled 100 Ohm differential line may require 
> that the single
> ended Zo will go up to 60+ Ohms (instead of 50 Ohms), but my 
> understanding
> is that the single ended Zo is not important as long as you 
> are terminated
> differentially.
> 
> Any opinions?
> 
> David Kaiser
> 310 Interlocken Pkwy.
> Broomfield, CO 80021
> (303) 460-4431
> david.kaiser@xxxxxxxxxx
> 
> ------------------------------------------------------------------
> 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: