a question:the accuracy of ads moment ?

when I use ads moment to simulate a lange coupler,the couple sames less than formulas caculates.but when I use HFSS simulate the same layout,the couple sames little bigger then theory.which one can I believe?

You should use probably a more dense mesh to simulate such problem in Mom( cells per wavelength>30, edge mesh on).

If you used adaptive meshing in HFSS and the convergence criteria was reached und most of all, you have correctly modeled, the results of HFSS will be more exact than that of MOM.

I do think the mesh density does'n affect this problem very much for I use a very dense mesh and there sames no improvement,also I found the jumper height maybe affect the couple of the lange.

Momentum does not accurately handle conductor thickness. This can
affect accuracy on edge coupled structures like Lange couplers. Your best bet is HFSS if you know your conductor thickness accurately.

i think HF%% is better.