It would be useful to know more about your set-up/sample.
I presume your excitation is ~330nm? the onset of a peak at ~650nm is therefore the second order of the excitation, and so the spectrum should be cut at ~600nm.
It would be useful to run an instrument response profile of, depending on your setup, something scattering but not luminescent.
Try to change Excitation and emission slits into 15 and 5, respectively. Also dulute the sample to get better readings. For more info, refer to the following link.