Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] Same geometry but slightly different results

Status
Not open for further replies.

waqsoh

Newbie level 5
Joined
Dec 23, 2009
Messages
8
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
pakistan
Activity points
1,392
i am doing HFSS simulations of simple microstrip. I designed two microstrips having same dimesnisons and everything, as shown in Fig 1. However when i see s-paramters for both microstrips, they are slightly different, as shown in fig 2. I dont understand why this happens. i tried denser mesh, but it still gives same result.
 

Attachments

  • twomicorstrips.PNG
    twomicorstrips.PNG
    46.8 KB · Views: 131
  • twomicorsparam.PNG
    twomicorsparam.PNG
    45.2 KB · Views: 136

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,522
Helped
1,597
Reputation
3,199
Reaction score
1,199
Trophy points
1,393
Activity points
35,079
Yes, is about meshing. The program choose a slightly different meshing for one of the structures (compared to the other one), and this gives slightly different results.
 

BigBoss

Advanced Member level 5
Joined
Nov 17, 2001
Messages
5,476
Helped
1,565
Reputation
3,132
Reaction score
1,464
Trophy points
1,393
Location
Türkiye
Activity points
33,014
It might be dependent also on Boundary Conditions.One boundary may be closer to one Microstrip or else..
HFSS uses Boundaries in EM Simulations that should be taken into account.
 

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,522
Helped
1,597
Reputation
3,199
Reaction score
1,199
Trophy points
1,393
Activity points
35,079
Could be the Boundary also, but if use: Draw/Region/Padding Type/Absolute Offset/Value, then the boundary will be at the same distance (which is "value") from any part of the structure.
 

waqsoh

Newbie level 5
Joined
Dec 23, 2009
Messages
8
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
pakistan
Activity points
1,392
It might be dependent also on Boundary Conditions.One boundary may be closer to one Microstrip or else..
HFSS uses Boundaries in EM Simulations that should be taken into account.
I have checked, all the boundary conditions are same. There is a question relating to boundaries. I have a microstrip enclosed in a vaccum box, with radiating boundary condition.
If i include a second microstrip enclosed in a separate vaccum box, with radiating obundary condition. Will these two microstrips have effect on each other, in terms of s-parameters.
In other words, having two designs in same hfss project and then having two designs in separate projects. Do these both schemes differ in terms of s-parameters.
 

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,522
Helped
1,597
Reputation
3,199
Reaction score
1,199
Trophy points
1,393
Activity points
35,079
Do these both schemes differ in terms of s-parameters.
Run the simulation and see the results. Hard to guess something like this.

Did you try using a Region with absolute offset instead of a Box, for the boundary?
 

waqsoh

Newbie level 5
Joined
Dec 23, 2009
Messages
8
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
pakistan
Activity points
1,392
Run the simulation and see the results. Hard to guess something like this.

Did you try using a Region with absolute offset instead of a Box, for the boundary?
I used region with absolute offset as shown in fig 1. S-parameters are shown in fig 2. There is still difference but it is small compared to previous case and at 6 GHz, sparams are almost same.
For comparison fig 3 shows the case when the microstrips are in separate boxes and the distance of each side from the box is same as the absolute offset case. Fig 4 shows the s-params for this case. In this case the s-params dont converge to same value.
 

Attachments

  • fig 1-design.PNG
    fig 1-design.PNG
    10 KB · Views: 103
  • fig 2-1-sparam.PNG
    fig 2-1-sparam.PNG
    48.1 KB · Views: 107
  • fig 3-design.PNG
    fig 3-design.PNG
    60.2 KB · Views: 106
  • fig 4-sparam.PNG
    fig 4-sparam.PNG
    46.4 KB · Views: 112

BigBoss

Advanced Member level 5
Joined
Nov 17, 2001
Messages
5,476
Helped
1,565
Reputation
3,132
Reaction score
1,464
Trophy points
1,393
Location
Türkiye
Activity points
33,014
Even the Microstrip Lines in separate Boxes Ports are Coupled in anyhow.Therefore you can not consider that "I separated and that's all" because Ports are still in interactions.This is RF propagation.Ideal curves are only and only obtained in case of infinite distance between the structures.
 

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,522
Helped
1,597
Reputation
3,199
Reaction score
1,199
Trophy points
1,393
Activity points
35,079
I think the measurement difference that you get now using Region is very good and better than any tolerance that can exist.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top