How do we subtract instrumental broadening from the experimental data to get the sample broadening? The prescription for that involves integral breadth and not FWHM which are not precisely the same and do not have the same values in general. So if I have to use FHWM instead of integral breadth, how does the formula gets modified (if at all)?

What would be the formula for subtracting the instrumental contribution if I have fitted XRD peaks with a pseudo-Voigt function?