AnsweredAssumed Answered

substrate definition -interface layers

Question asked by MESFET98 on Dec 6, 2003
Latest reply on Dec 8, 2003 by MESFET98

Like • Show 0 Likes0
Comment • 4

Hello, I am a bit confused ??? , how does momentum define the interface layers in the substrate definition? I though it was as follows:

If I have the following stack up:

Air
-------Strip cond2
Alumina2 Via via2
------- Strip cond1
Alumina1 Via via1
///////GND////////

and I define the thickness as follows:
Alumina1=100um
cond1=10um
Alumina2=100um
cond2=10um

how does momentum define the thickness
of Alumina2 layer? Is it from bottom of cond1 to bottom of cond2, or from top of cond1 to bottom of cond2? I thought it was from top of cond1 to bottom of cond2, but someone told me otherwise. Now I am not sure.

Depending how it is defined, the capacitance calculation will change.

If momentum defines this as from bottom of cond1 to bottom of cond2 then what I specified for alumina2 in this example will be actually 100um-10um=90um between cond1 and cond2.

If it specifies as from top of cond1 to bottom of cond2 then it will be as I specified in the example above.

Any clues?

No one else has this question

Outcomes

mfredrik Dec 8, 2003 1:57 AM

Momentum is using a technology which is called Method Of Moments. This implies that all metal conductors are simulated as infinite thin sheets of metal. What confuses you is the fact that ADS Momentum does have a way to specify the thickness of each strip. This thickness is only used for loss calculations, not during the actual EM simulation. Then a infinite thin sheet of metal is used.

There is an method using multiple conductor layers and vias to model the thickness of a conductor. The following documents does provide some more info:

Tech Info Session
Considering Metal Thickness in Momentum EM Simulations

Support Example
ADS MOM : Simulating lines with thickness

Support Document
How to model thick metal in Momentum

There will be a new feature in ADS 2003C (the upcoming release) which automates the somewhat tedious workaround described above.

Actions

MESFET98 Dec 8, 2003 4:04 AM

Based on the documents you suggested,

The thickness of the dielectric between the two cond layers in momentum is exactly as it is defined in momentum since there is no thickness associated with the cond layers (unless I manually define the thickness of the cond layer).

So if I specify Thickness=100um it is between the two metals no matter what the thickness of the metal is since they are infinitely thin. (Providing I don't define the thickness as the tech info session suggested)

correct?

Actions

glcr

Dec 8, 2003 3:01 PM

Yes you are correct! 8)

More specifically, the substrate definition file is where the physical layer stack-up & dimensions are defined.

And yes, this process will become smoother for thick metal loss calculations in the Jan release of ADS 2003C. :)

Actions

MESFET98 Dec 8, 2003 11:03 PM

Thank You. Now I see the light

Actions

4 Replies

mfredrik Dec 8, 2003 1:57 AM
Mark Correct
Correct Answer
Momentum is using a technology which is called Method Of Moments. This implies that all metal conductors are simulated as infinite thin sheets of metal. What confuses you is the fact that ADS Momentum does have a way to specify the thickness of each strip. This thickness is only used for loss calculations, not during the actual EM simulation. Then a infinite thin sheet of metal is used.

There is an method using multiple conductor layers and vias to model the thickness of a conductor. The following documents does provide some more info:

Tech Info Session
Considering Metal Thickness in Momentum EM Simulations

Support Example
ADS MOM : Simulating lines with thickness

Support Document
How to model thick metal in Momentum

There will be a new feature in ADS 2003C (the upcoming release) which automates the somewhat tedious workaround described above.
Like • Show 0 Likes0
Actions
MESFET98 Dec 8, 2003 4:04 AM
Mark Correct
Correct Answer
Based on the documents you suggested,

The thickness of the dielectric between the two cond layers in momentum is exactly as it is defined in momentum since there is no thickness associated with the cond layers (unless I manually define the thickness of the cond layer).

So if I specify Thickness=100um it is between the two metals no matter what the thickness of the metal is since they are infinitely thin. (Providing I don't define the thickness as the tech info session suggested)

correct?
Like • Show 0 Likes0
Actions
glcr Dec 8, 2003 3:01 PM
Mark Correct
Correct Answer
Yes you are correct! 8)

More specifically, the substrate definition file is where the physical layer stack-up & dimensions are defined.

And yes, this process will become smoother for thick metal loss calculations in the Jan release of ADS 2003C. :)
Like • Show 0 Likes0
Actions
MESFET98 Dec 8, 2003 11:03 PM
Mark Correct
Correct Answer
Thank You. Now I see the light
Like • Show 0 Likes0
Actions

substrate definition -interface layers

Outcomes

Related Content

Recommended Content