Questions about DGS simulation in AWR.

[fly2050]

I got the AWR DGS model from Manjunatha_hv's posts. https://www.edaboard.com/threads/106105/
Thanks for sharing by the way. I learn that awr could solve DGS problem nicely. But I'm still confused about some problems, could anyone help me?

1. The simulation of the AWR DGS project from Manjunatha_hv's posts is very time costing, is there any method to accelerate the simulation speed or AWR is slow for this kind of question?

2. Normally the DGS circuits is tested in open space, but the simulation is carried out in a closed structure, could this cause big difference between the simulation and the test results? And how to minimize the difference?

 

[Manjunatha_hv]

Hello,

1. Yes you can accelerate the simulation speed by setting the Grid value to 0.1 mm instead of 0.05 mm...
Also use the AWR new Axiem 3D planar simulator instead of EMSight to speed up the simulations...

2. Good, at least you have verified yes the Top Boundary condition needs to be OPEN or free space, I think i missed while sending this file...
But originally setup withopen space boundary option...


---manju---

 

[volker_muehlhaus]

2. Normally the DGS circuits is tested in open space, but the simulation is carried out in a closed structure, could this cause big difference between the simulation and the test results? And how to minimize the difference?

It causes a difference if your circuit has (much) radiation.
But another question: even if your filter is measured in open space, is it also used in open space later? Or do you have some box around the final product?

 

[fly2050]

Hi, manju

Thank you for your response. The Top Boundary of your model is OPEN, the Bottom Boundary is closed, is it ok or should all of the boundaries be OPEN? The simulation has ran 24 hours and just half of the frequencies been solved. Is it normal? Does it impact the accuracy when change the Grid setup?

---------- Post added at 08:17 ---------- Previous post was at 08:07 ----------

Hi, volker

I just simulate and test the circuit for research purpose, so I'm more concern about the differences between the simulation and the test results.

 

[Manjunatha_hv]

Hello,

Yes, all the TOP & BOTTOM Boundary conditions must be open...
its strange to see it is taking so long...Did u set the Grid to 0.05 mm & used Axiem 3D Planar MoM EM simulator or EMSight?
If you use Axiem it should be aroud 1 hr simulation time...

---manju---

 

[fly2050]

Hi, manju

I'v used both Axiem and EMSight to simulate. EMSight's speed could be accelerate by AFS. Axiem is very fast. But the results are different from each other. I think that's because the different boundry conditions of these simulators.
If I test the DGS in open space, the result of Axiem should be more accurate, am I correct? Thank you for your response by the way.

 

[Manjunatha_hv]

Hello,

Good to know that you have used both...But both results should match...
May be due to different boundary conditions...Yes, if you set Open space as the bondary conditions in Axiem you will get accurate results...
But remember to set the EM Edge Port properties Explicit Ground Reference option to as "Connetc to lower"...

---manju---