hfss wilkinson

I am modeling a 3dB wilkinson divider at 30 GHz. In ADS, the S parameters and performance is exactly what I would expect. I then export the layout in dxf and import it into HFSS. After adding all the geometries and ports to HFSS the model converges and solves. However, I have run into something that I can't explain. When I run the HFSS model with a lumped RLC (100 ohm) BC the S parameters are all screwy. When I remove the lumped RLC, the S parameters look just like a wilkinson but just a little wider than ADS.\

Any suggestions or comments?

Thanks

when u convert dxf, hv u check the unit is the same(mil/mm), one more, hv u add substrate and ground in hfss cos when u convert to dxf it just like 2-D mask

|
/ \
| |, :D

The units in ADS and HFSS are both set to um (including the HFSS dxf import screen). Also, the dielectric properties in HFSS are set to the same values in ADS. Any other suggestions?

Thanks for the reply

previously i import something from maxwell to HFSS using dxf techniques. But it was not hold good.

I feel the reason is for HFSS needs width of the metal strip, width of the substrate, . I talk about 3D.

Do you check your radiation boundary.....

Hi,

I have transferred many designs from ADS and with no problem. I have only imported the copper as dxf. Substrate, ports etc. have to added of course. Copper thickness in imported file dxf has been 0.1mm (set value) and it needs to be corrected. Have you checked that your substrate is attached to copper (no air between them)?
Have you placed your ports correctly with integration lines?

I can have a look at your file if you like.

Regards

Erno