a problem of ensemble

hi,i am designing the microstrip filer with ensemble and i had thought that the speed of ensemble is very fast but not very accurate till i found we can increase the mesh on the microstrip filter to
improve the accuracy but the time of simulating increase too ,and
i also cannt understand the"adaptive analysis"option in the "set up solution"option,i want to know how to use this option to improve the
accuracy of the filter!
thanks !

This is actually a good option to get an idea of convergence of result. Somtimes for a spacific mashing the simulated result may vary significantly if you go for more dense meshing. Using adaptive meshing you can have a converged result. But you may have to wait for longer time.

Many people things 20 celles/wavelength option in IE3D gives good results. But there is no way to check the convergence until you go for higher no of cells. In fact for coupld line structures 20 cell/wavelength gives large error.

Another thing is that a converged result may or may not be accurate. Accuracy depends upon the modelling. Say the edge feed in Ensemble does not use a good model.

fast sweep is a method of analysis that produces results faster not more accurate and requires you to set up appropriate bandwidth.
btw. in my experience ensemble produced accurate results (?) but i was analysing power dividers not filters :>

--
Rgz

hi,g86,i am designing a microstrip filter using HTS.the tranmission band is 2~2.025Ghz.when i use ensemble ,i donnt know whats the value in"adaptive frequency option"in"adaptive analysis option"is proper!and i found that if i change the value in"adaptive frequency option" ,the speed of convergency is also different.sometimes it cannot be convergent when i set the suggested value 0.05 in the"target maximum delta norm(q) option".why?and whats the value in "target.."is proper.
i need some information about these!
thanks in advance!