Ok, currently, how i detect water flow is by having a turbine inside the pipe that turns as water begins to flow. Now on the base of the turbine itself is a reflective and non-reflective layer. The problem is, the uP goes into LPM, but every 16ms wakes itself up, due to a routine that strobes the IR Tx to bounce off the turbine, which is then Rx'd by a Daylight filtered phototransistor. And thats what is my problem, currently the unit draws between 10-16uA, I want to try get that down to 5 or basically, only strobe the Tx once flow has been detected. The device should also be very VERY reasonable for this has to be low-cost. Yup, the wonderful life of utilities :)