Brain-computer interfaces can change the way that human body interacts with the external environment, making it possible to combine brain science technology with artificial intelligence for rehabilitation and diagnosis of brain diseases. Most of the related work reported so far is based on interface technology with wired connections, which can achieve multi-channel parallel EEG signals. However, the size and power consumption of these devices and systems are relatively large. Small, wireless, and low-power device systems are the key to achieve long-term, stable, and real-time data communication for brain-computer interfaces in natural environments. In this paper, we proposed a Field Programmable Gate Array (FPGA) based low-power data acquisition front-end solution design. Through the signal input of external patch electrodes, we combine the amplitude and frequency domain characteristics of in vivo brain electrophysiological signals to design a real-time integrated signal acquisition system, which is designed for real-time wireless transmission of scalp EEG signals through signal input from external electrode.