Multiplier and neural network synapse using current mirrors having low-power MOSFETs

摘要

A multiplier and a neural network synapse capable of removing nonlinear current use current mirror circuits The multiplier produces a linear current by using MOS transistors operating in the nonsaturation region. The multiplier comprises a first current mirror M1, M3, M7 to form a first current I 1 and a second current mirror M2, M4, M5 to form a second current I 2 , wherein the second current mirror is coupled in parallel to the first current mirror. The multiplier outputs an output current by subtracting the second current from first current, the output current representing the result of multiplying input voltages V 1 and V 2 . In a neural network synapse the input voltage V 1 is constant and the input voltage V 2 represents the synapse weight. An additional MOS transistor transfers the output current to the load in response to a neural state pulse.