I cannot get consistent answers when using the scipy.signal.welch function between the power spectrum density (PSD) 'density' and power spectrum 'spectrum' scaling options in Python 3.
I know that the units for density are V^2/Hz and and for spectrum the units should just be V^2. Therefore, I should be able multiple the PSD by the frequency resolution/bin bandwidth to convert to the power spectrum and vice versa, but I get inconsistent results. What detail am I missing.
