0
\$\begingroup\$

I'm trying to build a simple low-power water level alarm such that it could run continuously for at least 4 months on just 3x AA batteries. I've found this schematic which uses the 555 timer:

enter image description here

However, the 555 timer has a minimum power consumption of around 30mW, which would be too much. Would it be a good idea to modify the circuit such that the water's conduction supplies power to the 555 like so:

enter image description here

Is water conductive enough for this to operate?

Would it be a better idea to just use a low-power MCU like MSP430 and raise an interrupt when the probes short by water's conduction?

Any thought appreciated.

\$\endgroup\$
  • \$\begingroup\$ In the above schematic, load current will flow through the water as well. IMO, you should drive a transistor via the conductance of the water. \$\endgroup\$ – Rohat Kılıç Oct 12 '16 at 11:38
  • 1
    \$\begingroup\$ Many embedded processors can go into deep sleep. Some can drop down into the range of uA. However, experience says, creating an over all design that doesn't leak current can also be a challenge. \$\endgroup\$ – st2000 Oct 12 '16 at 11:42
  • \$\begingroup\$ the cmos version of the 555 has a power disapation of less than 1mW \$\endgroup\$ – JIm Dearden Oct 12 '16 at 11:50
  • \$\begingroup\$ @st2000 Could you elaborate on potential challenges? \$\endgroup\$ – John M. Oct 12 '16 at 11:51
  • 1
    \$\begingroup\$ (@st2000 Could you elaborate on potential challenges) When you get into the uA range you have to think your design through. Normally I wouldn't think twice about putting a 10K pull up on a N.C. switch to ground to pull the processor's pin up when the switch is activated. But think about it, when that switch is closed it draws 500uA. That washes away all the advantage of a, say, 10uA sleeping processor! Welcome to the world of portable battery powered devices. \$\endgroup\$ – st2000 Oct 12 '16 at 12:42
1
\$\begingroup\$

A few simple mods are all that is required. enter image description here

Add a p channel MOSFET to act as a ON/OFF switch for the 555. The circuit only draws current when the switch is activated.

The current through the MOSFET is small so anything that can carry 500 mA or more would be more than capable.

I've also simplified the oscillator to run from a single resistor (50% duty).

\$\endgroup\$
1
\$\begingroup\$

However enticing such a cheap circuit may be, don't try it. Long-term reliability is very, very bad. The problem is that the presence of any moisture in your sense path will cause corrosion to form a crust on your probe due to galvanic action. This will take some time under normal conditions, but it will happen. At this point the circuit will stop responding to water when is does arrive.

Any such sensor must have an AC sensing voltage, with no trace of DC.

\$\endgroup\$
  • \$\begingroup\$ A very important point +1 \$\endgroup\$ – JIm Dearden Oct 12 '16 at 14:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.