Ericson fluctuations in the differential cross sections were investigated for the compound reactions $^{27}\mathrm{Al}(p,{n}_{0})^{27}\mathrm{Si}$, $^{45}\mathrm{Sc}(p,{n}_{4})^{45}\mathrm{Ti}$, $^{45}\mathrm{Sc}(p,{n}_{5})^{45}\mathrm{Ti}$, $^{51}\mathrm{V}(p,{n}_{0})^{51}\mathrm{Cr}$, and $^{59}\mathrm{Co}(p,{n}_{0})^{59}\mathrm{Ni}$. Level widths in the compound nuclei $^{28}\mathrm{Si}$, $^{46}\mathrm{Ti}$, $^{52}\mathrm{Cr}$, and $^{60}\mathrm{Ni}$ were extracted from the analysis of the differential cross sections by the Fourier method and a nonlinear fit to the $\mathrm{ln}⟨{A}_{k}^{2}+{B}_{k}^{2}⟩$ distribution. The neutron spectra for each reaction were measured at least at three backward angles but the coherence width $⟨\ensuremath{\Gamma}⟩$ did not show a strong angular dependence. Nuclear level densities for the above nuclei were extracted by relating the average level spacing ${D}^{J\ensuremath{\pi}}$ to the average level width ${\ensuremath{\Gamma}}^{J\ensuremath{\pi}}$ using the fluctuation theory. Reasonable agreements were found with other level densities and level density compilations based in the Fermi gas formalism of Al-Quraishi and Huang for $^{28}\mathrm{Si}$, $^{46}\mathrm{Ti}$, and $^{52}\mathrm{Cr}$ but the comparison diverges for $^{60}\mathrm{Ni}$ for which additional measurements are required in order to clarify the observed discrepancies. The level density parameters that best describe the data are $a=3.5$ and $\ensuremath{\delta}=4.0$, $Ul25\phantom{\rule{0.3em}{0ex}}\mathrm{MeV}$, for $^{28}\mathrm{Si}$, $a=4.8$ and $\ensuremath{\delta}=\ensuremath{-}0.3$, $Ul20\phantom{\rule{0.3em}{0ex}}\mathrm{MeV}$, for $^{46}\mathrm{Ti}$, and $a=4.8$ and $\ensuremath{\delta}=0.3$, $Ul20\phantom{\rule{0.3em}{0ex}}\mathrm{MeV}$, for $^{52}\mathrm{Cr}$.