[SI-LIST] Re: Leaving, then re-entering a reference plane
- From: "Jason R. Miller" <Jason.R.Miller@xxxxxxx>
- To: silist <si-list@xxxxxxxxxxxxx>
- Date: Wed, 01 Jul 2009 12:16:44 -0400
Don,
If you do go down the path of simulating the split and looking at
what-ifs, there is a parameterized HFSS deck for simulating split planes
posted here:
http://www.hfss-forum.com/viewtopic.php?f=19&t=6
If you have any questions about the deck, feel free to contact me
directly or post them on the forum link above.
Thanks,
Jason Miller
Scott McMorrow wrote:
> Don
>
> Unfortunately, you have a classical non-localizable split plane problem
> that is not possible to understand with intuition alone. By
> non-localizable, I mean that the EM fields are not confined to a small
> region that can be easily analyzed. Non-localizable problems occur when
> there are plane excitations due to vias penetrating planes, or traces
> crossing plane splits. At the discontinuities, the signals will inject
> a parallel plate mode wave into the split core power plane, which will
> cause excitation at the 1st, 3rd and 5th harmonics of the incident bit
> pattern. When the pattern spectrum (and it's harmonics) aligns with a
> power plane resonance, a standing wave pattern will develop that will
> increase the power system noise, along with the signaling noise, jitter
> and crosstalk.
>
> What you will need to look for is the parallel resonant frequency of the
> power plane, that is, the resonant frequency of the bypass capacitor
> mounting inductance, and the plane capacitance. However, since there is
> no ground plane adjacent to the split power plane, the impedance will
> most likely be high on the average, and exceedingly high at resonance.
> This has the potential of being a very bad thing, and extremely
> sensitive to process variations of the PCB and the capacitors being used.
>
> You have 3 possible courses of action, in order of best to worst.
>
> 1) Redesign to remove the split.
> 2) Simulate it in a planar fullwave solver like Ansoft SIWave.
> 3) Build it and measure the signal and power system. A spectrum
> analyzer would be very useful to determine where the resonant peaks
> occur, and whether they will cause meaningful noise, crosstalk, or jitter.
>
>
> regards,
>
> Scott
>
>
------------------------------------------------------------------
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:
//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:
//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:
