Independent research centre, IMEC, has developed a battery-free wireless 2-channel EEG (electroencephalography or brain wave monitoring system) powered by a hybrid power supply using body heat and ambient light.

The hybrid power supply combines a thermoelectric generator that uses the heat dissipated from a person's head and silicon photovoltaic cells. The entire system is integrated into a device resembling headphones.

The system can provide more than 1mW on average when worn indoors, which is more than enough for the application. Thermoelectric generators using body heat typically show a drop in generated power when the ambient temperature is in range of the body temperature. Outside, the photovoltaic cells in the hybrid system counter this energy drop and ensure a continuous power generation. Moreover, they serve as part of the radiators for the thermoelectric generator, which are required to obtain high efficiency.

Potential applications are detection of imbalance between the two halves of the brain, detection of certain kinds of brain trauma and monitoring of brain activity.

The system is a demonstration of Holst Centre's Human++ program researching healthcare, lifestyle and sport applications of body area networks. Future research targets further reduction of the power consumption of the different system components of the body area network as well as a significant reduction of the production cost by using micromachining.

The thermoelectric generator is composed of six thermoelectric units made up from miniature commercial thermopiles. Each of the two radiators, on left and right sides of the head, has an external area of 4×8cm2 that is made of high-efficiency Si photovoltaic cells. Furthermore, thermally conductive comb-type structures (so-called thermal shunts) have been used to eliminate the thermal barrier between the skin and the thermopiles that is caused by the person's hair on the thermoelectric generator.

The EEG system uses IMEC's proprietary ultra-low-power biopotential readout ASIC to extract high-quality EEG signals with micro-power consumption. A low-power digital-signal processing block encodes the extracted EEG data, which are sent to a PC via a 2.4GHz wireless radio link. The whole system consumes only 0.8mW.