Mikhail,
You are likely correct. I'm not an RF engineer, so that's not my
bailiwick. Never the less, I'd like to analyze the PDN to quantify
the distribution of power noise to the RF circuits, so my problem remains.
Thanks!
Ralph
On 8/28/2015 9:23 AM, Mikhail Matusov wrote:
Ralph,
Spurs at -200 dB sound like complete nonsense. This is roughly thermal
noise of a 50 Ohm resistor in 10 Hz band at -273 C. And I haven't seen
an RF application where spurs at -120 dBc would be considered a
problem either.
/Mikhail
-----Original Message----- From: Ralph Wilson
Sent: Friday, August 28, 2015 9:40 AM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Looking for ideas......
All,
Some of my customers (in this case RF design engineers) are reporting
problems found
in their designs where noise on a power net that feeds the input to a
voltage
regulator (LDO, switching, or other) is conducted (but reduced in
magnitude)
to the output of the regulator. The output of the regulator supplies
power
(usually through subsequent L-C filtering) to an RF component. The noise
from
the upstream power feed reaches the RF component and shows up as spurious
noise on the RF signal. We're talking noise in the -120dB to -200dB
range, so these
are very sensitive RF nets. I would like to be able, in the simulation
world, to
see if I can come up with a quantitative analysis of the noise conducted
through
this power distribution network - in the most general sense of the
term, an "S21"
from the source power net to the client power pin(s).
In looking at some LDO data sheets, they spec a "ripple rejection rate"
(usually
around -40dB) specifically for this. I'm thinking I can model the LDO
(for AC purposes)
as a simply "Pi" circuit equivalent to the ripple rejection ratio of the
LDO - at
least as an initial approach.
One problem I am having is the tools (in this case Hyperlynx
SI/PI/Linesim)
don't support this kind of analysis. The power nets are typically
routed as area fills,
hence the trace modeler does not recognize them. Hyp PI sees the power
nets, but doesn't do S-parameter-like analysis (the Hyp AC wizard does do
Z-parameter analysis - I've been trying to think of how to bastardize
this to
get some meaningful results).
In theory I could go to 3D simulations, but this is not practical for
what I'm trying
to do. If I knew where the problem was on a specific design, I could
extract the
necessary pieces and probably measure what I already know. However, I'm
trying to
look at a fully routed design and analyze the whole PDN looking for
weaknesses
in the layout / routing - without any a-priori knowledge about whether a
problem
exists or where one might exist.
Right now I'm sort of stuck - I'm not coming up with any practical
means of
doing this analysis.
Looking for ideas.......anyone attempted this before - with success?
Which tool(s)?
Thanks!
Ralph Wilson
Alcatel-Lucent
------------------------------------------------------------------
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
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2015.0.6086 / Virus Database: 4409/10530 - Release Date:
08/28/15
