Mechanical polygon mirror (PM) based spectral filters provide a low-cost, alternative method of tuning through broadband spectra. Their application is predominantly in Swept Source Optical Coherence Tomography (SS-OCT) whereby large bandwidths and narrow laser linewidths are essential for producing high resolution images with adequate depth range. Whilst higher scanning speeds are possible from commercially available sources, PM spectral filters retain distinct advantages. They are versatile, capable of operating in any wavelength range and have the ability to run several sources from the same PM, thereby reducing their cost. These aspects make them suitable for experimenting with different settings for OCT imaging in the laboratory. Although a number of different experimental configurations have been reported, the exact method that achieves optimum performance still requires more research. This investigation examines a free space, PM spectral filter with a two-lens telescope arranged in a Littman configuration, utilising a transmission grating as the spectral dispersive element. Research has shown that the parameters of this optical system, such as the focal lengths used in the telescope, have a significant effect on the bandwidth and linewidth, as well as the maximum power throughput. The system’s overall performance can easily be diminished if careful attention is not given to their operation with respect to the other components. This study is a comparison of overall system performance between telescope designs of different lens numerical apertures and diameters and we examine the factors that lead to higher power transmission and more uniform spectral power distribution, by reducing aberration and vignetting effects. An analysis of the spectral power distribution carried out on 1-inch and 2-inch diameter telescope designs shows that the wider diameter telescope provides superior performance across the selected wavelength range.