Motor fault detection and diagnosis has been a core issue for many years as it is critical in maintaining industrial equipment. The integration of state-of-the-art signal processing and deep learning has been providing good promise recently in the performance improvement for such classification systems. This paper reports an investigation into the effect on the performance of Principal Component Analysis (PCA) and Wavelet Transform (WT)-based integration with a 1D Convolutional Neural Network (1DCNN)-based signal classification system. In fact, 1DCNN operates directly on the signal which allows it to be more efficient in terms of computational resources with less parameter requirements. In addition to this network, wavelet transformation was employed for extraction from raw signals, while PCA was used for dimensionality reduction. The processed data was given as input to the 1DCNN model. Wavelet transform makes it easier to extract the localized time-frequency information. Further, with PCA, redundancy and noise are reduced to much more robust features for CNN. The experimental results demonstrated the fact that PCA and wavelet transform would be able to help classification of motor faults by bettering the traditional methods. This integration was noted for capturing relevant features, which are crucial to classification, thus leading to an effective and accurate and fast model. The proposed method significantly the performance of systems for signal classification.