I have connected a Ketek AXAS SDD detector (Fig. 1) to an MCA-PRO 16 system and calibrated the spectrometer using a Fe-55 source (Fig. 2) purchased from Imagesco.


The resulting 55Fe calibration spectrum is shown on Fig. 3. 5.9 keV line width is 150 eV (2.5% FWHM).

57Co calibration spectrum is shown on Fig. 4. The 6.4 keV line is split and the 14.4 keV line appears right on the mark, confirming the expected near-perfect linearity of the silico drift detector.

Conclusion
57Co source is a better choice than 55Fe for calibrating silicon drift x-ray detectors. The 57Co source provides two far-spaced peaks (6.4 and 14.4 keV) that allow establishing / verifying calibration linearity over virtually the entire 1 to 20 keV range of the silicon drift detector (SDD).