[SI-LIST] Re: FW: Multiple power planes coupling to ground

Istvan and others,

Thank you for your insights. I have a better understanding of how the
planes will interact.

I agree that VCC2 will couple through the other power planes to ground
at high frequency, and this has 2 obvious side effects: 1) you will be
coupling VCC2 noise onto adjacent planes and 2) VCC2 is so far away
from GND, that it just won't filter well at high frequency.

Now I'd like to see just how useful  VCC2 will be and at what
frequencies will it stop being effective.

Calculating the capacitance, just considering 1 side of the case:

GND
|
VCC1
|
VCC2

If we just had VCC - GND with nothing in between, we'd use C = k (A/d)
to calculate the capacitance.

With VCC1 in there, VCC2's coupling to ground looks like 2 capacitors
in series. The value of these 2 capacitors is twice the capacitance of
what VCC2 - GND is, because they are half as far apart. However, since
we have twice the capacitance in series, we end up with the same
capacitance as we calculated from VCC-GND.

Now I think I can calculate the impedance of our VCC2 plane using the
Xc = 1/(2Pi*f*c) equation.

Does that make sense? Again, this discussion has been very helpful and
I thank you all for your inputs.

Joseph


>
>
> -----Original Message-----
> From: Istvan.Novak@xxxxxxx [mailto:Istvan.Novak@xxxxxxx]
> Sent: Monday, September 22, 2008 5:19 AM
> To: Joseph Pankow
> Cc: si-list@xxxxxxxxxxxxx
> Subject: Re: [SI-LIST] Multiple power planes coupling to ground
>
> Joseph,
>
> As always, it depends...  This stackup works fine as long as we make
> sure that the
> power planes are used primarily for low-frequency power distribution,
> where the
> capacitors attached to the planes will decouple the power planes from
> each other.
>
> At high frequencies VCC2 primarily couples to VCC1 and VCC3.  This is
> reciprocal,
> so noise on VCC3 is pushed on VCC1 and VCC3 and vice versa.  Also chances
> are that these plane layers are split, and in that case chances are that
> the splits do
> not line up vertically in the three power planes (if they did, why would
> we bother
> using separate planes?).  The plane shapes in any of the three power planes
> over splits will create bridges between the power domains also horizontally.
> So the structure at high frequencies may have a large number of cross
> couplings,
> making it more challenging to ensure that each power rail is properly
> bypassed.
>
> Regards,
>
> Istvan Novak
> SUN Microsystems
>
>
> Joseph Pankow wrote:
> > Hello all,
> > If you have multiple power planes sandwiched between 2 ground planes -
> > somewhere in your stackup - next to each other like so:
> >
> > |
> > GND
> > |
> > VCC1
> > |
> > VCC2
> > |
> > VCC3
> > |
> > GND
> > |
> >
> > From a high-frequency noise filtering/decoupling standpoint, VCC1 and VCC3
> > have relatively good capacitive coupling to ground, but what about VCC2?
> >
> > I understand that there will be capacitive coupling between VCC2 and the
> > other planes, but does this still give me a reasonably good filtering scheme
> > for VCC2 at high frequency? Or does it just couple VCC2 noise onto my other
> > power planes?
> >
> > I'm trying to visualize the loop currents to ground when devices draw power
> > from VCC2. Will the VCC2 power noise will couple directly through the other
> > power planes to ground, or will the other power planes will act like shields
> > so that VCC2 no longer has good low-inductance filtering for high frequency
> > noise?
> >
> > Any insight would be appreciated.
> >
> > Thanks,
> > Joseph
> >
> >
> >
>
------------------------------------------------------------------
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: