Thank you everyone for helping me with this situation.
I think what I stumbled over is the fact that in the
past I have always had a +5V plane, but in this case
the +5V is not so significant.
The best solution is probably to change the +5V plane
to +2.6V, and move the +5V somewhere else. I guess
it was just my old-timer bias. Am I getting to be a
dinosaur already? I'm kind of embarrassed....
Anyway, thanks again, I appreciate the help!
Jack
"Kowalewski,
Andy"
<AKowalewski@sych
ip.com>
To: "'Jack C. Olson'" <[log in to unmask]>
07/14/2003 04:23 cc:
PM
Subject: RE: [DC] Reference Planes for Impedance Controlled Signals
Caterpillar: Confidential Green Retain Until: 08/13/2003
Retention Category: G90 -
General
Matters/Administration
Either plane is going to be effective provided the planes are tightly
coupled (see my earlier notes). However you now have two planes, neither of
which is the ground plane, so the two power planes are now coupled only via
the the capacitance of each through the ground plane, and that's getting
tricky.
If it was my job, I'd make sure of it all by having at least one of those
planes next to the trace a ground plane, preferably closer than the power
plane you're talking about. Then you can be sure of at least one solid
return path, which is likely to swamp the impedance discontinuities as you
swap reference planes when the trace weaves in and out of each layer.
Not my idea of an ideal situation, but better than most.
And of course, running your signal trace over a split plane gets even more
hairy.
Bear in mind that all of this gets horribly worse as your rising and
falling
edges get shorter and shorter. You sure wouldn't be playing with variable
reference planes when transient times are less than 1nS - that's asking for
trouble. Signal integrity is likely to go out the window, literally. It'll
generate more EMI than the TV transmitter next door :-)
Andy K.
Sychip Inc
Office (972) 202 8852
-----Original Message-----
From: Jack C. Olson [mailto:[log in to unmask]]
Sent: Monday, July 14, 2003 2:19 PM
To: Kowalewski, Andy
Cc: '(Designers Council Forum)'
Subject: RE: [DC] Reference Planes for Impedance Controlled Signals
No, the traces are not imbedded.
The traces of interest are not on layer 1, they are on layer 4. But that is
not even important. Sorry if I was unclear.
The real question is:
What if the traces are interspersed between solid planes, and although they
are part of a +2.6V circuit, they happen to be closer to a +5V plane. I
think they will use the +5V plane but it doesn't matter.
Does it matter? Is a plane a plane?
"Kowalewski,
Andy"
<AKowalewski@sych
ip.com>
To: "'(Designers Council Forum)'"
<[log in to unmask]>
07/14/2003 02:04 "'Jack C. Olson'"
<[log in to unmask]>
PM cc:
Subject: RE: [DC] Reference
Planes for Impedance Controlled Signals
Caterpillar: Confidential Green Retain Until: 08/13/2003
Retention Category: G90 -
General
Matters/Administration
Need more info.
Is your trace of interest on layer 1 and the planes are in sequence below
that?
I assume you don't mean that the trace is embedded in the +2.6V or +3.3V
layers?
I'm not sure why you mention the 4th signal layer. Is that significant? Is
that where your trace of interest is?
The power and ground planes are equal reference planes if they are tightly
coupled together to AC signals. That happens if the planes are very close
together or there are lots of decoupling capacitors scattered around the
board that tightly couple them.
Of course, changing signal layers can change a microstrip configuration to
a
stripline, balanced or unbalanced, as well as change reference planes.
Andy K.
Sychip Inc
Office (972) 202 8852
-----Original Message-----
From: Jack C. Olson [mailto:[log in to unmask]]
Sent: Monday, July 14, 2003 12:56 PM
To: [log in to unmask]
Subject: [DC] Reference Planes for Impedance Controlled Signals
Reference Planes for Impedance Controlled Signals
I might be showing my ignorance by asking this question.
I thought I knew, but now I'm being challenged on it.
If I have impedance controlled signals that are part of
a circuit that is being run on +2.6V or +3.3V, but there
is a closer plane available, like +5V, the signal returns
will snap to the closer one, right? Does it matter which
TYPE of plane is being used a reference? The text I
received from a colleague implies that if my 2.6V plane
is too far from my 4th signal layer, which is next to the
+5V plane, then they will not be impedance controlled.
Doesn't sound right to me...
Jack (the new guy)
---------------------------------------------------------------------------------
DesignerCouncil Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF DesignerCouncil.
To set a vacation stop for delivery of DesignerCouncil send: SET DesignerCouncil NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site http://www.ipc.org/html/forum.htm for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
