The response of silicon-silicon-CsI(Tl) telescopes, developed within the FAZIA collaboration, to fragments produced in nuclear reactions 84Kr+120-124Sn at 35 A MeV, has been used to study ion identification methods. Two techniques are considered for the identification of the nuclear products in the silicon stages. The standard ΔE−E one requires signals induced in two detection layers by ions punching through the first one. Conversely, the digital Pulse Shape Analysis (PSA) allows the identification of ions stopped in the first silicon layer. The capabilities of these two identification methods have been compared for different mountings of the silicons, i.e. rear (particles entering through the low electric field side) or front (particles entering through the high electric field side) side injection. The ΔE−E identification method gives exactly the same results in both configurations. At variance, the pulse shape discrimination is very sensitive to the detector mounting. In case of rear side injection, the identification with the "energy vs. charge rise time" PSA method presents energy thresholds which are significantly lower than in the case of front side injection.