Serak,

Forgive me for jumping to conclusions, but it appears that you may be
attempting to calculate the impedance of a "differential pair", or what is
commonly sometimes referred to as "differential impedance", or more
correctly, an "impedance controlled differential pair". One common type of
signal that falls into this category would be LVDS, which has a
"differential impedance" of 100 ohms, but which could have each signal
independently routed where they were not in close proximity to each other
(no coupling) and each signal would then be routed as a 50 ohm line.

When the "differential pair" is routed side by side in close proximity to
one another such that there is intentional "coupling" between the two lines,
it will usually be in one of two structural forms. The first would be a
structure that consisted of two parallel traces over a ground plane, which
is called "differential microstrip", and the other would be a structure that
consisted of two parallel traces between ground planes which is called
"differential stripline". There is yet a third structure which consists of
two parallel traces where the traces are stacked one on top of the other,
both of which are between ground planes, and I have heard this referred to
as "broadside differential stripline", although I cannot vouch for the
accuracy of that terminology, nor do I know of any formula that can be used
to deal with this last type of structure.

The first two structures, "differential microstrip", and "differential
stripline", are actually quite common today, especially in routing LVDS
signals.

Unfortunately, I do not believe that Protel knows how to really deal with
either of these structures.

However I can recommend you to a calculator called "Zcalc", which is
designed to handle just such calculations. One of the nice things that I
like about "Zcalc", is that it will simultaneously calculate both "normal"
(single line) "impedance" and also "differential impedance" from the
dimensions that you provide, so that you can not only check the single line
"impedance" calculations against some industry standard such as "AppCAD"
from HP (Agilent), but you can also see and understand the relationships
between the different parts of the "structures" as you vary the different
dimensions within the "structure". "Zcalc" also has numerous links to other
valuable information built right into it which you can access just by
clicking the link in the lower right corner of the calculator, which will
take you to its webpage, which happens to be identical to the link below:

    ===> http://www.logiccell.com/~jean/LVDS/

The link that you want to download your own copy of "Zcalc" is the very
first link at the top of the page, with the date of 11/05/2001, and the
version number of 1.30b9.

A quick word of warning. There are many different calculators out there that
call themselves "differential impedance calculators", including some even
referenced to by "Zcalc" itself, that have an "error" built directly into
them, and therefore give incorrect answers. What happened here is that
National Semiconductor wrote the definitive paper on LVDS, the "LVDS Owners
Manual", and some other important related papers (some versions of which the
links in the "Zcalc" webpage will point you to), most of which are in PDF
Format. Well, the problem here is that there is an error in one of the
primary formulas given in the some of the PDF Literature, which appears to
be limited to the "Second Edition" of the one PDF Literature file mentioned
above ("The LVDS Owners Manual"), and it is not apparent from the PDF
Literature itself, and is only commented on and the error explained by the
National Semiconductor webpage that itself links to that PDF Literature.
Thus if you access the National Semiconductor PDF Literature from any place
besides the National Semiconductor webpage, you will never see the notice
about the error, and any calculator you build from just that version of the
PDF Literature itself will produce an error in the results. Well guess what?
Almost everybody gets to the PDF Literature by using some direct link
somewhere, and as a consequence goes directly to the PDF Literature, which
has the error. Oddly enough, even though "Zcalc" it self does not appear to
have the "error" internally, it still points you to the "First Edition" of
the PDF Literature (only 66 pages), and not the "Second Edition" of the PFF
Literature (101 pages). Possibly this means that the "First Edition" did not
have the "error" that the "Second Edition" does, and "Zcalc" was coded from
the "First Edition" without the error, and anything coded from the "Second
Edition" will have the error, unless one gets to the correct webpage with
the "errata". Anyway, the following link will not only get you to the
webpage that has the "errata" regarding the "errors" in the PDF Literatue at
the at the bottom of the webpage, but it will get you to the current "LVDS
Owners Manual" which is the "Second Edition".

  ====> http://www.national.com/appinfo/lvds/0,1798,100,00.html

Fortunately, since "Zcalc.exe" does not have the error, it can actually be
used to point out the error in some of the other "differential impedance
calculators".

There is some additional information in the PEDA archive that is quite
lengthy from the first time that we went thru this topic, and you should be
able to find it by simply searching for "differential impedance".

Hopefully I didn't "miss the boat" entirely as to what you were really
asking for, and hopefully this has all been of some help in answering your
question. Sorry that it wasn't much sooner, but I just today saw your
request.

JaMi

----- Original Message -----
From: <[EMAIL PROTECTED]>
To: "Protel EDA Forum" <[EMAIL PROTECTED]>
Sent: Friday, September 05, 2003 4:01 AM
Subject: [PEDA] signal integration


> Hallo,
>
> How can I find out imedance of track and capacity between 2 tracks?
> (I found some impedance in SI, but value 150ohm for 3mm x 10mm track is
nonsense)
>
> Serak
>



* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
*
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
*
* Contact the list manager:
* mailto:[EMAIL PROTECTED]
*
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
*
* Browse or Search previous postings:
* http://www.mail-archive.com/[EMAIL PROTECTED]
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to