[SI-LIST] Re: Matching Differntial Pairs

Content-Type: text/plain;
        charset="US-ASCII"
Content-Transfer-Encoding: quoted-printable
Kathy,

=20

=20

Fastest allowable PCI Express rise time is 50ps (see table below).

=20

=20

Good role of thumb is matching differential signals up to 1/10 rise/fall
time:

That means 5ps is about 28 mils matching.

=20

=20

Edi

=20

-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of Kathy Jacques
Sent: Thursday, June 21, 2007 10:50 AM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Matching Differntial Pairs

=20

Content-Type: text/plain;

      charset=3D"US-ASCII"

Content-Transfer-Encoding: 7bit

I am interested in some information regarding the routing of PCIe
signals.

I have worked with 2.5Gbps differential signals in the past, but I am
new to

PCIe.  I have noticed that PCIe requires add-in cards to have
differential

pairs matched to 5mils. =20

=20

=20

=20

Can anyone explain why the PCIe differential signals on an add-in card
need

to be matched with 5mil?  I have never had to match signals this closely
in

the past, so I wasn't sure of the rationale.  Is it for EMI reasons?  Is
it

really necessary to match this closely?

=20

=20

=20

If you are running the differential signal coupled (I have read that
this is

not necessary to route them as a coupled pair, but I have never
implemented

differential signals as uncoupled etch, and am reluctant to experiment
on my

current design - I can't afford to have the board not work because I
miss

some detail), then what is the best way to achieve this 5mil matching?
I

have seen several different techniques.  I have always been taught that
it

would be better to be mismatched by 30 mils and routed together the
whole

way, than to add etch to one signal to match the length.  If I do need
to

meet the 5 mils, and I need to add etch, then what is the preferred
method?

Certainly, adding the etch at the transceiver side would be preferable.

However, I don't have a good feel for HOW to add the etch.  I have
always

been told that serpentining (second pair below) is bad, but I see that
this

app note that I found is recommending the serpentine. =20

=20

=20

=20

Could someone please explain the preferred method for matching coupled

differential pairs, as well as why this method is better?

=20

=20

=20

Thank you for your help,

=20

kathy

=20

=20

=20

=20

=20

=20

=20

=20

=20

=20

=20

=20

-- Binary/unsupported file stripped by Ecartis --

-- Type: image/gif

-- File: image001.gif

=20

=20

------------------------------------------------------------------

To unsubscribe from si-list:

si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

=20

or to administer your membership from a web page, go to:

http://www.freelists.org/webpage/si-list

=20

For help:

si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

=20

=20

List technical documents are available at:

                http://www.si-list.net

=20

List archives are viewable at:    =20

            http://www.freelists.org/archives/si-list

or at our remote archives:

            http://groups.yahoo.com/group/si-list/messages

Old (prior to June 6, 2001) list archives are viewable at:

            http://www.qsl.net/wb6tpu

 =20

=20




-- Binary/unsupported file stripped by Ecartis --
-- Type: image/gif
-- File: image002.gif
-- Desc: image002.gif


------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List technical documents are available at:
                http://www.si-list.net

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu

Other related posts: