Re: water level sensor circuit for wireless monitoring
Hi,
You know what they say - Ask a vague question, get a useless answer...
Low power: 1mA or 1nA per day or per year? Is it in a dark enclosure? It it for 1,000,000 litres or 50ml? etc.
Anyway, the bulk of the current drawn is supplying the TLV431 with ~800uA. A reference with a lower supply current would be an idea, such as a resistive divider (but that would create a falling ref wrt battery voltage). The actual sensing circuit, which as you can see is just a variation on a very, very standard premise/design (the so-called "lie detector" aka "soil humidity sensor" aka "water level sensor" open or closed path to gnd or wherever the designer fancies), in simulations draws approximately 130uA to 170uA at 3V and will certainly work down to around 2.4V or lower. You could even dispense with the comparator, frankly but then have to spend time designing in battery voltage drop and ensuring Vbe and/or VOH is still attainable.
I'm interested in seeing serious suggestions for a low power water level sensing circuit.
- - - Updated - - -
Hi,
I'd thought of pressure sensor water level detecting too because I'd thought about avoiding corroding wires and so on, but at first glance it looks as though it is more complicated parts-wise and calculations-wise than a simple yes/no circuit really requires. e.g.
Liquid-Level Monitoring Using a Pressure Sensor
I simulated the above comparator circuit using a diode-connected NPN as the reference (instead of the TLV431) and it does use far less quiescent/housekeeping current and seemingly functions from 0º to 100ºC. Not sure I'd want to chance that sort of temperature-dependent reference unless I'd done some real testing first.
Far more economical re quiescent current is the bjt trio seen in so many of these circuits. Looking at what I was simulating, I think it's better to have the output go low when the water level drops below whatever level as that way the bulk of the current is drawn then and not all the time just to let the user know that there is water above x level. As you appear to be using a PIC/MCU that's not an issue I guess as input levels and what they signify can be defined in the code. Two more simulation screenshots for comparison of this idea about minimising current draw with active high or active low signals: