Thanks for the INP file. The error occurs because the FWHM of the peaks become large. When that happens the Simple Axial convolution complains. This is an exception that is deliberately thrown; I could in fact allow execution to processed but obviously something is wrong when peaks comprise 10s of millions of points.
The offending INP segment is:
prm pkz -0.00080_2.02715
prm pky 0.00001_2.02715
prm pkx 0.00207_2.02715
prm axial 0.00010_2.02715
prm u 0.00116_2.02715
prm v 0.00267_2.02715
prm w 0.00093_2.02715
TCHZ_Peak_Type(!u, 0.00116_2.02715,!v, 0.00267_2.02715,!w, 0.00093_2.02715,!pkz, -0.00080_2.02715,!pky, 0.00001_2.02715,!pkx, 0.00207_2.02715)
Simple_Axial_Model(!axial, 0.00010_2.02715)
Here we the pkz, pky etc... parameters defined twice, once within the TCHZ_Peak_Type macro and once using prm. Nothing wrong with defining them twice but in one definition they are refined and in another they are not refined. The parameters are actually refined depending on which prm statement is encountered first. If you dont want them refined then best to simply remove the prm lines, ie.
TCHZ_Peak_Type(!u, 0.00116_2.02715,!v, 0.00267_2.02715,!w, 0.00093_2.02715,!pkz, -0.00080_2.02715,!pky, 0.00001_2.02715,!pkx, 0.00207_2.02715)
Simple_Axial_Model(!axial, 0.00010_2.02715)
I think I may in the future stop execution if parameters are defined more than once and have different Refine attributes.
If you actually wanted to refine the peak shape parameters but want to limit the overall FWHM then you would need to put a Max limit in the TCHZ_Peak_Type macro. Note, the macro is typically well behaved but in your INP file the peak positions are starting way off their starting values leading to the peak shape parameters misbehaving. In any case to limit the FWHM size in the TCHZ_Peak_Type macro then modify the macro as follows:
pv_fwhm = Max(tch_p, LIMIT);
where LIMIT is some value, ie. something like 0.3 maybe.