cst frequency range setting

When simulate an antenna, I set different frequency ranges (0~3 and 0~6 GHz, resonant frequencies of the antenna are 0.9 and 2.4 GHz) for two transient analysis respectively. But I got very different results. In the simulations I have the same mesh properties. I can not figure out why CST can give so strange result. Would anyone pls explain me this or give some suggestions on how to select frequency range?
Thanks very much!

I really didn't understand your problem. You simulated one antenna for two different setups: 0-3 GHz and 0-6 GHz. Since you are telling that you got two very different results, let us consider: at frequency 0.9 GHz or 2.45 GHz you got different return loss for two different simulations; is it the case?
In my view it wouldn't be that different. If possible please upload your .cst file.
regards,
shameem

Did you try to use a finer meshing? Or the option "adaptive meshing"? - Most of the time meshing is critical...

Yes, the two results are not that different. The resonant frequencies are same, but there is 10 dB difference in S11. (the 0-3 setup is worse)
PS:
The mesh has 558000 meshcells, is it enough for a PIFA antenna with the size of 17*6*6 mm^3?

If you vary the number of meshcells without getting any change in S-paramaters it seems to be ok. - But maybe your design is just very sensitive.

Thanks! I will try ur method.