Recently infrared spectrometers are being used in industry, chemistry and medicine, agriculture and material sciences. These products are usually large and expensive. But recently small, inexpensive and mobile spectrometers are being produced and commercialized. In this study design of a visible near infrared spectrometer is considered and one of the usage areas are shown with an example application. HAMAMATSU c11708ma mini spectrometer sensor is used as optical sensor array. Spectral data which is collected using c11708ma sensor is transferred via UART to PC or STM3210e-Eval board. LDA is used as classification method. LDA model is created by PC application and transferred to STM3210e-eval board for further classification. Two different soil types are used to collect learning data. 1600 (800 for each) different soil spectrums are collected and used as learning data to create a LDA model. Classification success is %100 while using amplitude calibration.