Closed-loop feedback control for oxygen concentrator

Abstract

An oxygen concentrator is provided which has a first molecular sieve bed connected to a four-way valve (i.e., a cross-over valve) which either joins the sieve bed to a pressurized air source (compressed air) or alternatively vents it to atmosphere. A second molecular sieve bed is also joined to the four-way valve in a corresponding manner. The sieve beds are joined at the outlet end to a product reservoir. The sieve beds are also in fluid communication at the outlet end by a pressure equalization flow path. A concentration equalization valve regulates the flow in the pressure equalization flow path. In accordance with the present invention, the microprocessor uses a closed-loop feedback circuit to evaluate the amount of time that the output product gas is allowed to flow into the used sieve bed. Specifically, an incremental step advance is used assuming a peak value of the time of flow one direction as compared to the time of flow in the second direction. Thus, so long as the measured oxygen output to the patient keeps increasing, the control will direct incremental advances in the valve setting in the same direction. When the measured value decreases, the control steps back to the previous setting (i.e., the penultimate setting). In a second embodiment, control is based on a variable residence time range and variable pressure set point.