It works more or less as you described. The "water detect output" is pulsing in the simulation instead of being steady because {R7+RV1, C4} and (R6, C5) timing constants are very close. Instead, R7+RV1 shound be a combination that ranges between 3 and 4k Ohm, for a 250us timing of U1B while U1C would time about 400-500us. The first monostable U1B would give pulses when oscillation freq of U1A is low enough, while the second monostable U1C will output high if there are pulses, or low if there aren't.
The timing cap of the monostable made with U1D+U1E is not the 100nF one, but the 20uF. This monostable is working in "One Shot" more... Whatever happens to the input, it won't retrigger until reaching the intended elapsed time. The problem with this topology and the 40106 is that once U1E goes high and the capacitor has partially charged, then the input of U1D is going to see about 8V. This is dangerous because the "internal protection diodes at such input would be forwarded, and because of the internal design of these gates, a parasistic SCR could turn on and start a " latching Up" that would destroy the IC. The PNP transistor takes care of such situation. When U1D goes low, the emitter of the transistor will clamp the capacitor voltage to about 4.5V. Because the PNP isn't very fast in this case, the 100nF cap acts as a filter so the possible voltage spike never reaches more than 5V. Because the 20uF capacitor now (dis)charges below Vcc, there is a voltage divider at U1E output to prevent that a low level would lead the input of U1D below 0V. I prefer to not use a zener diode at U1D input because the 20uF (multilayer ceramic) is charging very slowly due to the 1/10th duty time pulse generator in order to get about 60s of timing (in that time the mini pump would extract about 2L from the tank).
The circuit has been proven to work nice in breaboard, and now i'm porting it to a protoboard for installing in an enclosure that i've yet to model for 3D printing.
The sensor is hand made. It's just a pair of high endurance enameled wire of 2m length, weaved, interleaved, in a plastic form of 2.6*3cm printed, ironically, in PET filament made of drinking water bottles. The sensor, when dry, has 85pF of capacity, when soaked in distilled water, about 1050pF, when pulled from the water while still humid, It has 650-700pF. Adjusting the trimmer for detecting a rise of 850pF is working fine.
Home
Help
Search
Login
Register
