Yahoo Groups archive

Homebrew PCBs

Index last updated: 2026-04-05 19:38 UTC

Message

Re: [Homebrew_PCBs] Re: Another rookie starts a high volume pcb project.

2005-07-14 by David Cureton

Hi Stefan,
    It is my understanding that the copper islands or orphaned copper is 
removed to help with  EMC  (or do they call it FCC in the US) approval.

You are probably right in that a fat copper island will not do any harm. 
However at very high frequencies
these isolated (in a DC sense) pieces of copper may just have the right 
shape to form L's and C's that become resonant
and hey - presto , you got yourself an antenna.

Although the chances of  a floating piece of copper  being resonant is 
probably pretty slim, when designing for EMC you take all the free runs 
that you can get

Also - When I say very high frequencies -  the signal cycle times do not 
have to be high frequencys - they just need fast transition times.


Cheers,
David




Stefan Trethan wrote:

>On Thu, 14 Jul 2005 09:55:36 +0200, Phil <phil1960us@...> wrote:
>
>  
>
>>However, I wouldn't focus on using fat traces for avoiding copper
>>removal.  I think that would be the hard way to do it.  I much prefer
>>to create a ground polygon that encompasses the entire board and then
>>let the layout software figure out the actual shape.  In eagle, I put
>>a ground poly on both the top and bottom with isolate set to 24 mil.
>>Then I route the board.  The only places eagle will take copper is for
>>isolation and "orphaned" copper areas (not connected to ground).
>>Look at http://www.geocities.com/pcbs4less/boardtop.gif for an example
>>of this. Red and green are copper areas/traces.  As you can see, this
>>design keeps much more copper than one using just fat traces.  It
>>seems much cleaner and makes for a quieter board, noise-wise, because
>>of the extensive ground areas.
>>Besides, thin traces just look more professional to me
>>Phil
>>    
>>
>
>
>I agree, reasonably thin looks better.
>I also make groud planes, but i always wondered about islands. You say you  
>don't leave the copper there, which i only partially understand (though i  
>do the same, i just don't know exactly why).
>
>How bad are they? I mean i can imagine they could potentially provide more  
>capacitive coupling between signals which is unwanted, but is that really  
>an issue?
>Have you any information about that?
>
>Sometimes those gaps don't look right, and i'm tempted to leave them.
>
>I usually use 10mil for signal traces and 20 mil for power, but i've also  
>made boards with 1mm traces for customers if they want, where 10mil would  
>have been plenty, and it doesn't look bad if the components are right for  
>it (only throughhole, not many ICs, ..) Esp. if all traces are that wide  
>and none are thinner it looks ok.
>
>ST
>
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBs
>
>If Files or Photos are running short of space, post them here:
>http://groups.yahoo.com/group/Homebrew_PCBs_Archives/ 
>Yahoo! Groups Links
>
>
>
> 
>
>
>
>
> 
> ** ACCEPT: CRM114 PASS Markovian Matcher ** 
>CLASSIFY succeeds; success probability: 1.0000  pR: 75.3676
>Best match to file #0 (nonspam.css) prob: 1.0000  pR: 75.3676  
>Total features in input file: 7075
>#0 (nonspam.css): features: 1444288, hits: 289711, prob: 1.00e+00, pR:  75.37 
>#1 (spam.css): features: 554608, hits: 88467, prob: 4.29e-76, pR: -75.37 
> 
>
>-=-Extra Stuff-=-
>
>  From sentto-4505361-9835-1121335815-yahoo=dcureton.com@... Thu Jul 14 20:10:50 2005
>Return-path: <sentto-4505361-9835-1121335815-yahoo=dcureton.com@...>
>Envelope-to: yahoo@...
>Delivery-date: Thu, 14 Jul 2005 20:10:50 +1000
>Received: from n10.bulk.dcn.yahoo.com ([216.155.201.63])
>	by whitetail with smtp (Exim 4.34)
>	id 1Dt0g5-0004wZ-6G
>	for yahoo@...; Thu, 14 Jul 2005 20:10:49 +1000
>Comment: DomainKeys? See http://antispam.yahoo.com/domainkeys
>DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=lima; d=yahoogroups.com;
>	b=Lb4DZCCehu3oAy78nYfZeu+7BDSrOiZZrHZQ5hCwm4hl4xridI2icJdqPe3ctklw2D4hKWCqeQr/ezqKR2yZdMhfQ83Vb6EH43qSLCDE1Mn8ldxv4QsahkJpW9jJf9Yo;
>Received: from [216.155.201.65] by n10.bulk.dcn.yahoo.com with NNFMP; 14 Jul 2005 10:10:15 -0000
>Received: from [66.218.69.3] by mailer2.bulk.dcn.yahoo.com with NNFMP; 14 Jul 2005 10:10:15 -0000
>Received: from [66.218.66.29] by mailer3.bulk.scd.yahoo.com with NNFMP; 14 Jul 2005 10:10:15 -0000
>X-Yahoo-Newman-Property: groups-email
>X-Sender: stefan_trethan@...
>X-Apparently-To: Homebrew_PCBs@yahoogroups.com
>Received: (qmail 22942 invoked from network); 14 Jul 2005 10:10:14 -0000
>Received: from unknown (66.218.66.216)
>  by m23.grp.scd.yahoo.com with QMQP; 14 Jul 2005 10:10:14 -0000
>Received: from unknown (HELO mail.gmx.net) (213.165.64.20)
>  by mta1.grp.scd.yahoo.com with SMTP; 14 Jul 2005 10:10:13 -0000
>Received: (qmail invoked by alias); 14 Jul 2005 10:08:01 -0000
>Received: from v211-056.vps.tuwien.ac.at (EHLO tu-x2pj5qeyp2u4) [128.131.211.56]
>  by mail.gmx.net (mp008) with SMTP; 14 Jul 2005 12:08:01 +0200
>X-Authenticated: #9664799
>To: Homebrew_PCBs@yahoogroups.com
>References: <db55po+ei0u@...>
>Message-ID: <op.stwefmsbmg0lsf@tu-x2pj5qeyp2u4>
>In-Reply-To: <db55po+ei0u@...>
>User-Agent: Opera M2/8.01 (Win32, build 7642)
>X-Y-GMX-Trusted: 0
>>X-eGroups-Msg-Info: 1:12:0
> From: "Stefan Trethan" <stefan_trethan@...>
>X-Yahoo-Profile: stefan_trethan
>Sender: Homebrew_PCBs@yahoogroups.com
>MIME-Version: 1.0
>Mailing-List: list Homebrew_PCBs@yahoogroups.com; contact Homebrew_PCBs-owner@yahoogroups.com
>Delivered-To: mailing list Homebrew_PCBs@yahoogroups.com
>List-Id: <Homebrew_PCBs.yahoogroups.com>
>Precedence: bulk
>List-Unsubscribe: <mailto:Homebrew_PCBs-unsubscribe@yahoogroups.com>
>Date: Thu, 14 Jul 2005 12:08:00 +0200
>Subject: Re: [Homebrew_PCBs] Re: Another rookie starts a high volume pcb project.
>Reply-To: Homebrew_PCBs@yahoogroups.com
>Content-Type: text/plain; charset=iso-8859-15
>Content-Transfer-Encoding: 7bit
>
>On Thu, 14 Jul 2005 09:55:36 +0200, Phil <phil1960us@...> wrote:
>
>  
>
>>However, I wouldn't focus on using fat traces for avoiding copper
>>removal.  I think that would be the hard way to do it.  I much prefer
>>to create a ground polygon that encompasses the entire board and then
>>let the layout software figure out the actual shape.  In eagle, I put
>>a ground poly on both the top and bottom with isolate set to 24 mil.
>>Then I route the board.  The only places eagle will take copper is for
>>isolation and "orphaned" copper areas (not connected to ground).
>>Look at http://www.geocities.com/pcbs4less/boardtop.gif for an example
>>of this. Red and green are copper areas/traces.  As you can see, this
>>design keeps much more copper than one using just fat traces.  It
>>seems much cleaner and makes for a quieter board, noise-wise, because
>>of the extensive ground areas.
>>Besides, thin traces just look more professional to me
>>Phil
>>    
>>
>
>
>I agree, reasonably thin looks better.
>I also make groud planes, but i always wondered about islands. You say you  
>don't leave the copper there, which i only partially understand (though i  
>do the same, i just don't know exactly why).
>
>How bad are they? I mean i can imagine they could potentially provide more  
>capacitive coupling between signals which is unwanted, but is that really  
>an issue?
>Have you any information about that?
>
>Sometimes those gaps don't look right, and i'm tempted to leave them.
>
>I usually use 10mil for signal traces and 20 mil for power, but i've also  
>made boards with 1mm traces for customers if they want, where 10mil would  
>have been plenty, and it doesn't look bad if the components are right for  
>it (only throughhole, not many ICs, ..) Esp. if all traces are that wide  
>and none are thinner it looks ok.
>
>ST
>
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBs
>
>If Files or Photos are running short of space, post them here:
>http://groups.yahoo.com/group/Homebrew_PCBs_Archives/ 
>Yahoo! Groups Links
>
>
>
> 
>
>
>
>
> 
>  
>  
> -0-0-0- 
>  
>

Attachments

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.