I have problem with Rietveld refinement using GSAS_EXPGUI software. When I try a refinement diffraction pattern obtained from analysis of a known powder (as SiO2, CeO2) I have a big problem in relative intensities of the peaks. In fact, my peaks have a relative intensity different from the reference pattern. Also, after phase parameter refinement I don't see an improvement, and my chi square is very high (200-300). Where is my error? I prepared the powder milling with agate mortar, and after that I put the powder in a specimen holder filling the hole. If I try the same refinement process with a diffraction pattern coming from the network (as a tutorial or example pattern) I haven’t had problems and obtain good results.