I assume that you are getting too much signal. In this case look for a different spectroscopic line for the element that has a lower response. If it is a matter of too little signal and too much electronic noise, then look for a stronger spectroscopic line or increase the path length or find a way to concentrate the sample, or any of the other methods for getting higher signal.

If the elemental lamp in your spectrometer is surely well-adjusted, try increasing the amount of light in the optics and on the detector by increasing the supply current (HCL lamp) or power (EDL lamp) and also by increasing the width of the monochromator slits. But it may cause an excessive drop in measuring sensitivity. If this does not improve the situation, use a different lamp or a different spectrometer.

Sometime ago, I observed high gain with while optimizing As line in our GF-AAS. I noticed that as the lamp gets weaker, the gain increases. With a new As lamp, the gain is lower. Regardless, As (193 nm) usually have high gain, presumably due to the line. More so, changing the line to a higher wavelength may affect the sensitivity. Therefore, ensuring that the lifetime of the lamp is not exceeded, adequate lamp optimization, and proper line selection might help get the needed lamp gain to work with.