From: S. Weir ([email protected])
Date: Thu May 10 2001 - 12:10:10 PDT
I don't think it's a good idea to propagate rules without an understanding
of where and why they work. A mitered corner is not going to prevent
tandem traces, any more than a square corner will force them. The same
holds true for edge coupling between adjacent traces on the same
layer. However, the "tandem", and "colinear" rules in Allegro do address
these issues directly.
If we return to Dr. Johnson's book he refers to the frequencies where
mitering actually starts to show benefits. This is reinforced by Mr.
Swanson's PDF, that implies the difference converges somewhere below 200pS
rise times. From the data, it doesn't look to be too much of a concern
until we start to sink below 100pS, which only some boards are facing today.
At 02:15 PM 5/10/01 -0400, you wrote:
>The argument is that this Zo (and the resulting voltage amplitude) dip for
>typical today's digital systems is just too small to worry about. But this
>'game' of using right-angle bends if used irresponsibly may cause havoc in
>a high-speed printed circuit board. The cause of havoc is not from the
>corner itself, but from what the corner gives rise to.
>If you have two signal layers on a printed circuit board where the tracks
>are routed orthogonal to each other (and there is no plane separating the
>two signal layers), right-angle corners can eliminate the orthogonality
>and cause tracks (in the two layers) to run parallel to each other -
>probably at long lengths - and give rise to cross-talk problems. In
>today's dense boards, it's not uncommon to see such signal layers
>separated by 5 mils or less. With tracks running parallel and separated by
>that much, and worse still, coupling in a broadside sense (as opposed to
>edge coupling), unforeseen crosstalk problems can occur. And not just
>crosstalk problems, a tight coupling of that sort also gives rise to an
>impedance discontinuity causing further signal amplitude loss due to
>So, I see this great distaste for right-angle bends, as just like those
>childhood myths that are basically untrue, but nevertheless providing a
>useful end result.
>>From: Dagostino, Tom [mailto:[email protected]]
>>Sent: Thursday, May 10, 2001 1:09 PM
>>To: 'Chris Padilla'; [email protected]
>>Subject: RE: [SI-LIST] : MECL System Design Handbook
>>If you look at the cross section of the trace at the bend it gets
>>wider. A wider trace has more C per unit length. More C will look like
>>a lower Zo. I've seen dips in impedance TDR'ing right angles.
>>IBIS and Tau Modeling Manager
>>Mentor Graphics Corp.
>>From: Chris Padilla [<mailto:[email protected]>mailto:[email protected]]
>>Sent: Thursday, May 10, 2001 9:46 AM
>>To: [email protected]
>>Subject: Re: [SI-LIST] : MECL System Design Handbook
>>It is not a myth, there is "some kind of effect" from right angle bends.
>>The real question is, I believe, "Do I care, given the speed and/or
>>frequency content of the signals I am dealing with?"
>>For most of us right now, on this list, we probably *should not* care as
>>you point out.
>>If one is designing a 50 GHz antenna, then it might make a world of
>>Out of curiosity, what kind of rise time were you pushing through that
>>right angle bend? 100 ps? I am farily certain that I've seen some small
>>dips (aka capacitance) through right angle bends but I cannot honestly say
>>that it was directly attributed to the bend.
>> >Any of you who want to know how the myth about right angle bends got
>> >started, look
>> >at figure 7.17 on page 155. This alleges that right angle bends can be
>> >seen by a
>> >TDR. I've done this measurement dozens of times and coiuld never see a
>> >angle bend.
>> >A few years ago, I called Mr. Blood the author of the book and asked
>> >about the
>> >diagram. His reply was that he knew the diagram was flawed, but there
>> >wasn't time
>> >to fix it before the book went to press.
>> >As a result, thousands of engineers have spend countless time worrying
>> >about right
>> >angle bends.
>> >When we publish technical information such as this, it is important to
>> >insure it
>> >is accurate.
>> >This applies especially to applications notes, whic often contain entirely
>>**** To unsubscribe from si-list or si-list-digest: send e-mail to
>>[email protected] In the BODY of message put: UNSUBSCRIBE
>>si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
>>si-list archives are accessible
**** To unsubscribe from si-list or si-list-digest: send e-mail to
[email protected] In the BODY of message put: UNSUBSCRIBE
si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
This archive was generated by hypermail 2b29 : Thu Jun 21 2001 - 10:11:54 PDT