Gears are one of the most important elements in industry for motion and force transmission. Condition Based Monitoring for gears became one of the essential fault detection fields, especially in early condition. An approach of fault detection in frequency domain will be introduced. The FFT power spectrum will applied to the vibration signal come from gear box after some signal processing analysis for accurate decision. The vibration signal was exposed to an adaptive filter FFTFilt which his coefficients extracted from Linear Predictor Coefficient (LPC). Nonlinear Energy Operator (NEO) is used to magnify the vibration signal with high energy or sudden energy and decrease the signal contain low energy. For online monitoring as well as real-time analysis NI LabVIEW was used to implemented the suggested analysis algorithm The gear fault were detected in more than one speed, and comparison between vibration signals in frequency domain shows the importance of proposed algorithm.