Sorry to make this so long,

Basically, controlled impedance gives everyone some sort of
idea about how signals are going to react to a "known" or
"given" condition.  This given condition is known as impedance.
A fancy name for resistance. In a given condition, one would
know that a signal of 5 volts peak at a frequency of 300 million
cycles per second (otherwise known as Hertz or Hz) passed
along a 50 ohm trace for 3 inches into another device,
such and such will happen AT LEAST ON PAPER. Depending
upon alot of factors, this given impedance condition may
be life or death to the signal or the information it
contains.  For instance, in some CATV equipment I was
involved with the FCC had laws for levels of signals out
on the telephone pole miles from the Central Office. The
performance of this equipment couldn't be predicted unless
the parameters of the cable were known and matched somehow
to all(most) of the parameters of the board attached to
the cable.

Process factors:  with FR-4 could vary performance from one
side of the board to the other.

Laminates: Only Fr-4 experience.

Line Width: Sorry to do this to you but ...

  Impedance (Z) is determined by the formula

  Z0 = squareroot(L0/C0) where

  L0 = inductance per length
  C0 = capacitance per length.

  So this Z0 is impedance per length (ohms per length).

  Inductance is the ability to store and release electrical
  energy from current.

  Capacitance is the ability to store and release electrical
  energy from voltage.

  In other words, impedance is a measure of how efficient
  a trace will carry a signal with electrical energy. The
  higher the impedance, the less efficient it is.

  Increase your trace width, and you are effectively making
  a bigger capacitor between the trace and the ground plane.
  Look at the equation above... increase C0 and you DECREASE
  Z0 by the square root of c0.  Double the width of your
  trace which is the same as doubling the area of a plate
  for a capacitor and you decrease your Z0 by 0.707.

Transmission Frequency: Increase the frequency and there
  there will be a point reached (I wrote a small paper on
  this - it gets really complicated) that is basically the
  resonant frequency of the trace construction. Below this
  point the response of the trace is essentially flat, i.e.
  no loss in signal integrity (in an ideal world of course).
  Above this point and the response of the trace degrades
  at a -20dB per decade slope, i.e. it begins to degrade
  pretty fast. In an ideal case, my little analysis came
  up with about 850 MHz for FR-4.

              Resonant Point
    ------------
                 \
                   \
     Trace Response  \
                       \
                         \
    +----------------------
      Frequency


Regards,  Doug


************************************************************

----------
From: <David Albin> <[log in to unmask]>
To: [log in to unmask]
Subject: [TN] Definition of controlled impedance ?