The impact of the magnetic field profile on the operations of plasma-based Electron Cyclotron Resonance Ion Sources (ECRIS) has been debated and investigated for a long time, and a summary is given by the collection of the so-called “magnetic scaling laws”, i.e., some rules-of-thumb providing the optimal configuration of the axial and radial confining magnetic fields. Anyway, the optimal configurations have been found just looking at the output beam currents and charge states. Only recently, more advanced experiments have shown that the optimal performances are related to some “islands" in the parameters space leading to a dense, energetic but stable plasma. A general, systematic characterization of plasma density and temperature as a function of the magnetic scaling, and for different energy domains (i.e., for cold, warm, and hot electrons) is not yet available. This paper presents a multidiagnostics characterization of a hydrogen plasma heated by microwaves at around 6.8 GHz via Electron Cyclotron Resonance, and confined in an axisymmetric simple mirror magnetic trap. Investigation of plasma response to solely the axial field profile gives the unique opportunity to decouple the effect of simple mirror field from the radial components that are, usually, provided by a hexapole. Optical emission spectroscopy, Langmuir probe, and an x-ray silicon drift detector were simultaneously used to evaluate the plasma density and temperature in different energy domains, thus exploring the electron energy distribution function from a few eV to hundreds of keV. Plasma parameters have been measured as a function of the magnetic field profile (in particular by varying the ratio B_{min}/B_{ECR}), of the microwave power and of the neutral pressure. Furthermore, for the same scaling, the relative abundances of H atoms and H_{2} have been estimated. Data show smooth trends of electron density and temperatures versus microwave power and neutral pressure, but strong and even nonlinear responses as a function of the magnetic field profile variation, allowing us to connect semiempirical scaling laws to changes in plasma parameters.