Hi Guys,
Bill is correct in that it is the rise time rather than the frequency that
is of importance. However, the frequency does determine the maximum
bandwidth.
Rule of Thumb: All drivers whose trace length (in inches) is equal to or
greater than the rise time (in ns) must have provision for termination. So
for a 1ns rise time the max length is approx one inch.
Also, the bandwidth of a 40MHz signal needs to be considered. I generally
take the 5th harmonic (or 7th if it is extremely fast rise time). So 40MHz
means you are really working with 200MHz.
For differential pairs - they are best routed with a clearance of twice the
trace width. So, for a 4 mil trace the clearance should be 8 mils.
There are quite a few articles, on these subjects, that were originally
published in the PCB Design Magazine on my web site if you go to
www.icd.com.au and then "News and App Notes" from the main menu.
Practical Signal Integrity
Differential Pair Routing
Etc
Cheers,
Barry
-----Original Message-----
From: DesignerCouncil [mailto:[log in to unmask]] On Behalf Of Brooks,
William
Sent: Tuesday, 24 September 2013 3:25 AM
To: [log in to unmask]
Subject: Re: [DC] Impedance Control vs. Line Length
Hey Jack,
Is this RS232? Ethernet? Some other mode?
Impedance matching is important... especially as you get into higher
frequencies...
At Microwave frequencies, the reflections that happen at the transition from
one impedance value to another can destroy the integrity of the signal...
40Mhz sinusodal analog signal has a wavelength of about 24 1/2 feet (7.5
meters), (1/4 wave length = 6.125 ft).
A 40Mhz digital signal has frequency components that are in the Gigahertz to
make the square wave signal... (that's a little different animal) ....
The rise/fall time of the digital signal is key... A 1 GHz has a wavelength
just under a foot... (0.98357 Ft) 1/4wave would be about 3 inches...
That means that a 3 inch line on a pcb will make a great 1/4 wave antenna
for 1Ghz signals... a 1.5 inch line works well for 2 Ghz... etc... receiving
or transmitting...
So check your rise times... they are the component that sets the high freq
for your signal.... not the clock rate. The mutual close coupling of
differential traces can help reduce x-talk due to the nature of
differentially transmitted signals... there are tradoffs to know about in
routing differential traces... The more tightly coupled the more sensitive
to changes in the trace geometries... the looser they are coupled the less
sensitive to routing influences.. Lee Ritchey talks about all of the aspects
of high speed signal integrity in his 'Right the first time' book...
http://www.thehighspeeddesignbook.com/ He did a presentation for the
Designers Council local chapter here in San Diego awhile back and it was
very informative...
My reccomendation would be to run the lines differentially matched to the
required impedance for the signal you are trying to route regardless of how
short they are. Better safe than sorry... ;)
Best regards.
William Brooks, CID+
Senior MTS (Contract)
2747 Loker Ave West
Carlsbad, CA 92010-6603
760-930-7212
Fax: 760.918.8332
Mobile: 760.216.0170
E-mail: [log in to unmask]
-----Original Message-----
From: DesignerCouncil [mailto:[log in to unmask]] On Behalf Of Jack
Olson
Sent: Monday, September 23, 2013 8:26 AM
To: [log in to unmask]
Subject: [DC] Impedance Control vs. Line Length
I haven't done much impedance control lately, and I can't remember at what
length the impedance control becomes significant.
Is it 1/4 wavelength?
1/10 wavelength?
1/20 wavelength?
In other words, if I have very short traces compared to the frequency,
should I even worry about it?
(I have some Tx/Rx pairs that theoretically should be routed as differential
pairs, but they only need to be routed about an inch away. The clock is
40Mhz)
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
----------------------------------------------------------------------------
-----
DesignerCouncil Mail List provided as a free service by IPC using LISTSERV
16.0.
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF DesignerCouncil.
To temporarily stop/(restart) delivery of DesignerCouncil send: SET
DesignerCouncil NOMAIL/(MAIL) For additional information, or contact Keach
Sasamori at [log in to unmask] or 847-615-7100 ext.2815
----------------------------------------------------------------------------
-----
____________________________________________________________________________
_
Scanned by IBM Email Security Management Services powered by MessageLabs.
For more information please visit http://www.ers.ibm.com
This email is intended only for the use of the party to which it is
addressed and may contain information that is privileged, confidential, or
protected by law. If you are not the intended recipient you are hereby
notified that any dissemination, copying or distribution of the email or its
contents is strictly prohibited. If you have received this message in
error, please notify us immediately, by replying to the message and deleting
it from your computer.
WARNING: Internet communications are not assured to be secure or clear of
inaccuracies as information could be intercepted, corrupted, lost,
destroyed, arrive late or incomplete, or contain viruses. Therefore, we do
not accept responsibility for any errors or omissions that are present in
this email, or any attachment, that have arisen as a result of e-mail
transmission.
____________________________________________________________________________
_
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
----------------------------------------------------------------------------
-----
DesignerCouncil Mail List provided as a free service by IPC using LISTSERV
16.0.
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF DesignerCouncil.
To temporarily stop/(restart) delivery of DesignerCouncil send: SET
DesignerCouncil NOMAIL/(MAIL)
For additional information, or contact Keach Sasamori at [log in to unmask] or
847-615-7100 ext.2815
----------------------------------------------------------------------------
-----
______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask]
______________________________________________________________________
---------------------------------------------------------------------------------
DesignerCouncil Mail List provided as a free service by IPC using LISTSERV 16.0.
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF DesignerCouncil.
To temporarily stop/(restart) delivery of DesignerCouncil send: SET DesignerCouncil NOMAIL/(MAIL)
For additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
---------------------------------------------------------------------------------
|
