Mail*Link(r) SMTP               FWD>Controlled Impedance - Dielectric Constant

Avner-

The dielectric constant Dk for insulative materials is just that -- (in
theory) a constant.  

In reality, the Dk of most materials varies with something, be it frequency,
temperature, humidity, etc.  In the case of printed board base materials, the
Dk for epoxy-glass (e-glass) FR-4/GE&GF-like dielectric materials, the epoxy
varies in Dk with frequency, for all practical purposes the e-glass doesn't. 
The Dk for epoxy-(e-glass) varies with frequency from about 4.7 @ 1 MHz to
about 4.3 @> 100 MHz.

The Dk used for striplines calculations (center conductor with two outer
"signal ground planes" will be the materials Dk at the frequency of concern,
or you will need to select a number based on a band/range of frequencies.

In the case of microstrip transmission lines (a signal conductor separated
from the signal ground plane), an adjustment must be made for the Dk to
compensate for the dielectric constant of the air over the conductor (and any
other stray capacitances).  Generally, the most microstrip transmission line
calculation use "sqrt(Dk + sqrt(2))" for the composite or "effective" Dk.

In the case of "thin" dielectric coatings (such as solder resist) over
microstrip transmission lines some other factors must be included to
compensate for the addition insulative material and the effect on microstrip
impedance.  As a rule-of-thumb the we use, (based on the work of Kaupp) he
determined that "thin" (less than one conductor thick, Dk ~ 4.5 ) base
material coatings over the microstrip's signal conductor reduced the impedance
about 5%.  "Thicker" (greater than 2 conductor thicknesses) coating reduced
the impedance by as much as 20%.

Humidity is another factor and can be a problem in critical microstrip
transmission lines.

If you have low manufacturing toleranced "critical" transmission line
impedances you've got to quantify the process and materials.

For brevity, co-planar wave guides will be skipped for now.

Ralph Hersey
e-mail:  [log in to unmask]
  
--------------------------------------
Date: 5/21/96 6:12 AM
From: Eltek Ltd.

Subject: Controlled Impedance - Dielectric Constant

I like to know if anyone found that Dielectric Constant, for the same
material (e.g. FR-4),
 might be different when used for calculation of Microstrip or for
calculation of Stripline.

Avner Drory 


------------------ RFC822 Header Follows ------------------
Received: by quickmail.llnl.gov with SMTP;21 May 1996 06:11:06 -0700
Received: from ipc.org by simon.ipc.org via SMTP (940816.SGI.8.6.9/940406.SGI)
	 id IAA25224; Tue, 21 May 1996 08:07:57 -0700
Resent-Date: Tue, 21 May 1996 08:07:57 -0700
Received: by ipc.org (Smail3.1.28.1 #2)
	id m0uLqdw-0000CjC; Tue, 21 May 96 07:34 CDT
Resent-Sender: [log in to unmask]
Old-Return-Path: <[log in to unmask]>
Date: Tue, 21 May 1996 15:42:16 +0300
Message-Id: <[log in to unmask]>
X-Sender: [log in to unmask]
X-Mailer: Windows Eudora Light Version 1.5.2
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: [log in to unmask]
From: "Eltek Ltd." <[log in to unmask]>
Subject: Controlled Impedance - Dielectric Constant
Resent-Message-ID: <"uvrqY2.0.BhC.BXRen"@ipc>
Resent-From: [log in to unmask]
X-Mailing-List: <[log in to unmask]> archive/latest/4240
X-Loop: [log in to unmask]
Precedence: list
Resent-Sender: [log in to unmask]