Hey guys,

I'm making an alarm that produces regular timed beeps whenever its probes are connected by water (well, strictly speaking coffee, not water). I have a working prototype now, based on a two transistor flasher circuit, but I want to use a 555 to time the beeping instead. In my simulation, wiring the 555 to the MOSFET source seems to work well, but I can't test it the other way around (load on drain), because the application uses an implicit ground for the 555. Anyway, do you think wiring the 555 to drain will work better?

I've attached the circuit as it is now. The 30k and switch models the coffee liquid. The pots adjust frequency and duty cycle. The 2.8k is a piezo beeper.

 

Hey guys,

I'm making an alarm that produces regular timed beeps whenever its probes are connected by water (well, strictly speaking coffee, not water). I have a working prototype now, based on a two transistor flasher circuit, but I want to use a 555 to time the beeping instead. In my simulation, wiring the 555 to the MOSFET source seems to work well, but I can't test it the other way around (load on drain), because the application uses an implicit ground for the 555. Anyway, do you think wiring the 555 to drain will work better?

I've attached the circuit as it is now. The 30k and switch models the coffee liquid. The pots adjust frequency and duty cycle. The 2.8k is a piezo beeper.

Have you considered leaving the 555 powered all the time, and controlling it via the reset pin?
Also, do you know that if you pass DC through submerged contacts, they will corrode? You can avoid this by passing AC through them, but the circuitry is more complex.

 

I have considered leaving it on, but that would mean either drawing power all the time, or requiring a switch. I prefer to avoid both. But I should probably explain what it's supposed to do in more detail:
One probe clips to the edge of my aluminium coffee pot, and the other probe end is suspended half a centimeter or so above the floor of the chamber that fills up with coffee when it's finished. Such that current flows from one probe, through the aluminium walls, through the coffee, and to the other probe. When the alarm sounds, I will usually get up and remove the device (and pour my coffee) immediately, so current only flows between the probes for, say, half a minute or so. Do you think that would be sufficient to produce troublesome corrosion?

 

I would change the power switch to a logic level PMOS (Google), and drive the beeper directly from the 555, if possible.
I don't know how fast corrosion occurs, but keep in mind you will be getting some metal ions in your coffee. Don't use an arsenic electrode.

EDIT: The anode corrodes. No matter what kind of switch you use, this will be your probe. I'm not sure what material you should use. It appears that all metals are toxic at some level. Maybe carbon would be a good choice.