Substrate Noise Analysis

I need to run some substrate noise analysis to determine the cross talk
that will occur on an analog cmos circuit. I noticed that Cadence has
two software packages I could use, either SeimsIC or SubstrateStorm.
I'm wondering if someone could tell me what the advantages are to each
of them and which one I should be using for testing the crosstalk on my
analog circuit.

Thanks,

Scott McLean

 

[tattvamasi@gmail.com]

SeismIC - Approximate - Used for larger designs/circuits
Substrate storm - Rigorous, only to be used for smaller
designs/circuits.
one can also extracte the substrate and bolt it and then run a closed
loop simulation with extracted RC's and the substrate.

Both these tools involve setting up of a lot of technology files..
Substrate storm needs a complete doping profile from technology from
where the device modelling leaves off, till a critical distance of the
substrate.

We have used Substrate storm to a reasonable success in detecting
coupling between our PA and our LNA, but YMMV.

You would need assistance from Cadence setting up these files.

I hear that CAdence has discontinued these tools, and offers Assura-RF,
bundling up a subset of Substrate storm utils in Assura.

Someone from Cadence can offer more accurate info

Partha

 

[fogh]

scott.mclean@gmail.com wrote:

I need to run some substrate noise analysis to determine the cross talk
that will occur on an analog cmos circuit. I noticed that Cadence has
two software packages I could use, either SeimsIC or SubstrateStorm.
I'm wondering if someone could tell me what the advantages are to each
of them and which one I should be using for testing the crosstalk on my
analog circuit.

Thanks,

Scott McLean

For an analog circuit, you should be able to do this from assura RCX. Coupling
through whatever capacitive path (conductors or epi/well), and subtracte
resistance, are part of the extracted parasitics.

 

[Gerry Vandevalk]

Be aware that the current Cadence tools do ignore inductance ( A long time
complaint of mine)
(no pun intended!)

If you are looking at frequencies less that 5Ghz and not too long, you can
probably ignore the
fact that the substrate is lossy. ( The tools you mentioned do tend to
ignore this!)

Are you looking for specific coupling or just generic noise due to coupling?

Also what level of noise/coupling can you tolerate?

-- Gerry




<scott.mclean@gmail.com> wrote in message
news:1124981792.722226.35620@g14g2000cwa.googlegroups.com...

I need to run some substrate noise analysis to determine the cross talk
that will occur on an analog cmos circuit. I noticed that Cadence has
two software packages I could use, either SeimsIC or SubstrateStorm.
I'm wondering if someone could tell me what the advantages are to each
of them and which one I should be using for testing the crosstalk on my
analog circuit.

Thanks,

Scott McLean