Hi Istvan,
Yes, I agree with you on the model usability frequency range and possible
model breakdown. Just a few comments. The reason for the breakdown of the
impedance description in case of PDNs are mostly ports (size and type) and
model simplification (2D approximation of the Maxwell's equations).
The impedance matrix or operator in general do not have any frequency or
dimensional limitations and usable up to optical frequencies, that is very
similar to the scattering parameters for all practical cases (see R.F.
Harrington or R.E. Collin for instance). Though, to maintain the accuracy,
the ports either should be much smaller than the min wavelength or wave
ports should be used - this is in addition to 3D formulation for the
discontinuities. Also, open, short or any other termination conditions (for
s-parameters) are either equally accurate or all break down in similar way
as the ports become non-conservative (electrically large). Even port
de-embedding does not help in such cases. However, scaling ports down or
modal decomposition at the port helps to extend the accuracy of the
multiport formulation up to very high frequencies. It all depends on the
problem formulation and model construction and not the final descriptor.
Though, as you also mentioned, at high frequencies the analysis of PDN is
actually not relevant to power delivery and mostly needed for EMI/EMC and
signal integrity evaluation.
Best regards,
Yuriy
Yuriy Shlepnev, Ph.D.
President, Simberian Inc.
2629 Townsgate Rd., Suite #235, Westlake Village, CA 91361, USA
Office +1-702-876-2882; Fax +1-702-482-7903
Cell +1-206-409-2368; Virtual +1-408-627-7706
Skype: shlepnev
www.simberian.com
Simbeor - Accurate, Productive and Cost-Effective Electromagnetic Signal
Integrity Software
2010 and 2011 DesignVision Award Winner, 2015 Best In Design&Test Finalist
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On
Behalf Of Istvan Novak
Sent: Thursday, October 13, 2016 5:24 AM
To: shlepnev@xxxxxxxxxxxxx; thomas.beneken@xxxxxxxxxxx;
si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: PDN simulation - Trans-impedance versus
self-impedance
Yuriy,
Further to your point, I think the frequency range of interest makes a big
difference. In the frequency range where there the noise on the
board can get to the silicon through the power path, the impedance
approximation is quite valid. Up to a few MHz or few times ten MHz
open is open enough, and the PDN approximation of silicons as either open
or current source should also be a fair approximation on decent PDNs.
At higher frequencies the approximation breaks down, but those noise
components will mostly stay on the board and can screw up signal
integrity and/or EMC.
Regards,
Istvan Novak
Oracle
On 10/12/2016 10:30 AM, Yuriy Shlepnev wrote:
Just a few comments to the answer from Thomas.multiports and PDN characterization.
1) The impedance matrix is a complete description of unterminated
multiport system. It is so called open-circuit matrix that is obtained
with the following experiment. We apply unit current source to one of
the ports and measure voltage responses at this and all other ports,
that are open-circuited (unterminated). The voltage at the port with
the source is the self-impedance and the voltages at the other ports
are the off-diagonal elements of the impedance matrix or
trans-impedances. They can be used to compare qualitatively PDN
configurations - zero trans-impedance, for instance, means ideal
isolation of two power delivery systems (more relevant for EMC/EMI).
System with lower trans-impedances will have better isolation of the
components - though, even in this case properly terminated
S-parameters would be preferable. Also, to evaluate the impedance at
the component location or possible supply voltage variations, we need
to terminate all ports of the multiport (by normalization or reduction
of all ports to one). The trans-impedances are always in the mix as
parts of the diagonal impedance matrix elements. To understand it
better, you may need to read some introductory books on theory of
impedances are significant.
2) Both may be misleading, because of absence of port terminations.
Best regards,
Yuriy
Yuriy Shlepnev, Ph.D.
President, Simberian Inc.
2629 Townsgate Rd., Suite #235, Westlake Village, CA 91361, USA Office
+1-702-876-2882; Fax +1-702-482-7903 Cell +1-206-409-2368; Virtual
+1-408-627-7706
Skype: shlepnev
www.simberian.com
Simbeor - Accurate, Productive and Cost-Effective Electromagnetic
Signal Integrity Software
2010 and 2011 DesignVision Award Winner, 2015 Best In Design&Test
Finalist
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Thomas Beneken
Sent: Wednesday, October 12, 2016 3:23 AM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: PDN simulation - Trans-impedance versus
self-impedance
Hi Liviu-Dumitru,
1) a variation of current into pin 9 translates with a factor of
Z[2,9] into a voltage ripple on pin 2
2) the voltage ripple on a pin adds up from self produced current
variations (self impedance) AND current variations on other pins of
the PDN (transfer impedance). So the answer is self and transfer
trans-impedance ?
Hope this helps.
Thomas
Am 12.10.2016 um 11:37 schrieb Craciun, Liviu-Dumitru:
Hi all,course, in the "real world" there are more IC power pins.
for the PDN simulation I placed the Port[1] at the output pin of the
VRM (voltage regulator module) and the Port[2], Port[3], ... Port[11]
at the IC power pins.
The PDN with only 10 IC power pins is only a simplified example, of
The power integrity simulation software is able to calculate thethe VRM and each IC power pin.
self-impedances Z[1,1] at the VRM and Z[2,2], ... Z[11,11] at the IC.
There are the diagonal elements of the impedance matrix.
The elements Z[1,2], Z[1,3], ... Z[1,11] are the trans-impedances
between
In generally, the PDN requirements in the IC data sheet contain only"trans"-impedance.
ONE parameter - Z value @ Freq. - without to specification "self"- or
two IC power pins ? For example Z[2,9]
MY QUESTIONS:
1) What is the physical significance of the trans-impedance between
mean, self- or trans-impedance?" The standard answer is "only the
2) The IC provider can `t answer the question "what impedance do you
impedance". So the discussion with the IC provider is a "dead end
street". I am confuse ... what impedance is significant, self- or
Best regards,
Liviu Craciun
Harman Becker Automotive Systems GmbH Karlsbad, Germany
P.S.
I expected from the IC provider TWO impedances as specifications,
self- and trans-impedance, But I received ONLY ONE value. It is
unknown if the spec means the self- or the trans-impedance.
------------------------------------------------------------------
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 forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
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 forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
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 forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
