Learning Forum

[wx5]

I am using AnsysEM HFSS to do the signal integrity project. To define the materials effective dielectric constant and loss tangent, I created two datasets. I entered the value of dielectric constant and loss tangent from 1GHz to 10GHz as shown in Fig1. My simulation is from 0GHz to 64GHz.

 

I also tested if dielectric constant and loss tangent is constant after 10GHz, shown in Fig.2. However, the results shown a large variance, shown in Fig3.

 

My next hypothesis is that dielectric constant and loss tangent will be linearly increasing/decreasing after 10GHz as shown in Fig4. However, this simulation took more than two days and HFSS crushed during the simulation.

 

What will be the behavior of the dielectric constant and loss tangent after 10GHz?

 

Fig1. Dielectric constant(left) and loss tangent(right)

Fig2. Dielectric constant and loss tangent constant after 10GHz.

 

Fig3. Simulated results (red: only three data points entered from 1GHz to 10GHz; blue: Dielectric constant and loss tangent constant after 10GHz)

 

Fig4. Dielectric constant and loss tangent will be linearly increasing/decreasing after 10GHz

[Praneeth]

Hi,

Are trying to see how HFSS is extrapolating the material properties beyond the defined range?
Please let us know the type of material and application that you are using with this dataset to help us serve you better.

All the very best,
Praneeth.

[wx5]

Hi,

Yes. I do not know what HFSS is assuming after the material properties beyond the defined frequency range. I picked a random dielectric material as my substrate. However, when I defined the dielectric properties at a certain frequency, for example, beyond 10GHz, dielectric constant is 2.5 and tanD is 0.02, the simulation cannot converge. Thank you.

Best,

Wenting

[Praneeth]

Hi Wenting,

Please let us know how you are using this dataset in the material propeties window. Also share the screenshot of the convergence to guide you better. Did you increase the passes and see if it converges?

Best regards,
Praneeth.