Why are battery power-supply, fiber data transfer and small front-end size so important ?   The combination of a battery-powered front-end and optical fiber offers minimal leakage currents in both normal operation an under fault conditions. During normal operation this assures the best possible rejection of all kinds of interference. Common mode as well as differential mode interference voltages are therefore reduced far below the amplifier noise level. Although high CMRR may solve the common mode problem in mains powered systems, the differential mode interference remains at least a factor of 10 worse than with our battery powered setup. An additional advantage of the battery powered setup is the very high protection for shock-hazard caused by the mains-supply. Because of the fiber optic data transfer, the safety situation is completely determined by the capacitance between the subject and the environment. Because the dimensions of the front-end are very small compared to the dimensions of the subject and because the front-end is located very close to the subject, the capacitance between the subject and the environment is not significantly increased by connecting the front-end to the subject. Therefore the biopotential measurement does not alter the situation in terms of protection against shock hazards and leakage currents are well below the IEC 601-1 CF specifications.   Leakage currents through Zdrl cause Common mode interference (Vcm) AND Differential mode interference. The first one can be eliminated with a high CMRR, but the last one can not be decreased by the instrumentation. The amount of leakage current is directly dependent on the value of the isolation capacitance. The best way to keep this capacitance as small as possible is the use of a compact battery powered front-end close to the subject, in combination with fiber-optic data transfer to the data processing hardware.       Battery powered   Mains powered     What happens when multiple AD-boxes are connected to 1 subject?   Connecting two A3 systems to the same subject is not a good measurement setup. There will be errors when channels of one system are referenced to channels of the other system. There are two reasons for these errors. Firstly, the channels are measured with respect to different references (each system has its own CMS and DRL). Secondly, the internal clocks of the systems are not synchronized, so there will be slight timing differences between the sample moments of the two systems, which leads to "sampling skew" distortion. The above errors do not play a role during BioSemi Triggered Hyperscanning with multiple subjects, because in that case channels of one subject are not referenced to channels of another subject.