See attached picture. Would this circuit work to read water level? I'm using (I think) an AC voltage dividing system. One resistor being the water itself, the other being the fixed resistor. On either side of the fixed resistor I rectify the signal to go into a microprocessor. In theory would this circuit work?
Using ac signal so as to not crud up the probes/water which a DC signal will do over time.
UsernameD:
See attached picture.
Attached pic?
Nothing attached, but anything using conductivity will foul, and the electrodes will degrade over time.
Other options include float switches and noncontact methods such as capacitive sensors, ultrasonic sensors and pressure sensors.
That’s not a voltage divider, the ac is source is directly connected to the bridge rectifier input.
No matter as that kind of design will kill someone. Time for another approach.
OP’s schematic:
Oh boy!....if the bridge rec and mcu wasn't there, then can we agree that the circuit is a voltage divider since the water is acting as one of the resistors?
Also, the ac signal would be approximately 6vac, not 110vac...no ouch ouch.
Voltage doesn’t kill, current does. One milliamp is the human safe level and the trip limit for GFCI breakers. Water makes any current dangerous as it reduces the resistance and increases the contact area of skin. It’s just a really bad idea but more importantly, it’s an unworkable solution since the conductivity of water can and will vary. You can detect the presence of water, the level, not so much.
https://www.asc.ohio-state.edu/physics/p616/safety/fatal_current.html
Regardless of the electrodes, the circuit is setup to read Vac and will not be affected by conduction in the water (within reason).
Regarding the Vac source. If it mains voltage or connected to the mains in any way it is lethal.
A rule of thumb regarding mains. If you have to ask a question you should not be using it directly.
Watts. ..while agree that current kills, not voltage, the 6volts is not enough potential to cause harm.....again...no ouch ouch. Anyway.....with that put behind us....John, your point is that since it is transformed from 120vac to 6vac, this is not ok? I find this extremely hard to believe. Are you sure that standards such as etl or ul would not approve? I would think that they would since it runs through a transformer? If your saying to sense through a hall sensor or coil induction to completely isolate from the vac then I can understand, but wouldn't the transformer "pass" for safety as well?
No matter, AC or DC, 6V or whatever V, conductivity probes are NOT a way to measure water level. You can merely measure water presence, and even that fails for pure water (as that's non conductive).
For water level measurements there are many options: ultrasound from above or below, pressure, float switches, maybe even a capacitive sensor can work. Conduction is not one of them.
The variation in conductivity between different samples of water spans many orders of magnitude.
For very pure water it varies significantly with temperature.
Conductivity measurements are only useful for measuring the conductivity of the water sample.
Any use of conductivity to measure level would need to be ratiometric.
MarkT:
The variation in conductivity between different samples of water spans many orders of magnitude.
Indeed - there's about 5 orders of magnitude between soft tap water and sea water. Makes building an EC meter that can handle that complete range a bit of a challenge (been there done that, got a range of almost 6 orders of magnitude - the 2-3 orders of magnitude needed for general hydroponics were a bit easier :-)).
Mark and Marle both very good points! Marle I tested different water sources and you precisely right on range of water conductivity. I tested with different water and I found bottled water to read very low (about 1.5v passed via voltage divider circuit) and almost 7v via a known high salinity well water (same vd circuit).
Mark, right....this is not a suitable way to read water level, but you can use it as a way to dispense water, given you have a beginning reference value which would therefore let you know when to stop dispensing. You just have to know how much water volume is in between to top and bottom of your probes.
Well... still not with a simple voltage divider.
The ions in the water migrate, so when you send DC through the probes they migrate and the resistance you measure goes up (conductivity goes down) quickly. On top of that you get serious corrosion on your probes unless you use very resistant metals.
The proper way to measure conductivity of water is by applying AC, in the 1 kHz to 1 MHz frequency. And that's what makes it so hard.
WattsThat:
One milliamp is the human safe level and the trip limit for GFCI breakers.
No it isn't.
Hi,
Is you project aiming to show how much water has been dispensed, or to give a signal when the water reaches a certain level?
Can you explain the application?
What material is the container/tank?
What sort of volume of water is being measured?
Thanks.. Tom.. 
