The 10k resistors are too much off a load for the equivalent resistance of the sensor.

I would first measure the sensors resistance. From there, pick a pull up or pull down resistor so that you have a clear transition of logic.

Let's say your sensor has a resistance of 500k. And you want active high. So a pull down resistance of 4x of that would approach the St pins input threshold (80 percent Vcc). I would go above that and pick a 5x or 6x multiplier.

I think those types of resistance. Will expose to a slew of issues. As such, inpendence translation near the sensors is the way to go.

but how to measure sensor resistance. My sensor are electrodes which are suspended in the tank. Are u suggesting that when two sensors are short due to water level, that resistance to be measured? say if the resistance is 500k , so 5x will be 2.5 Mega ohm... correct?

 

You can measure the resistance with a multimeter.

 

Hi, my name is swapnil. I am trying to make a water level controller using arduino. I refered many circuit ideas for different website and made my own design. Simulated it in Protues and it worked fine. Also i designed my PCB in eagle and started implementing.
Everything was going as per plan, but when i tried to test the input sensors voltage, i got major voltage drop due to water acting as resistance between the sensors.
For sensor - I used a PVC pipe, in which Cat 6 LAN cable is inserted. One wire of LAN cable is connected at bottom of PVC pipe with a stainless steel metal. This wire is also connected to 5V supply through board. Same way other sensors are connected to different levels of PVC pipes. Now when water rises till 1/4 tank full sensor, arduino will get the voltage at its digital input pins (Because of water, 5V sensor and 1/4 tank full sensor will get short).
Now the problem here is, water is conductive but with high resistance since its not much salty, so instead of getting 5V at 1/4 tank full sensor, i am getting around 2-3 volts.
My board is also ready and sensor set for both Below tank and overhead tank are both ready, and my program also works fine. but i am stuck at this voltage drop problem.
I had one thing in mind, what if i use a intermidiate circuit betweent he sensor set and board, which will detect any voltage above 0V and give output 5V. But i am not getting any such circuit ideas to build one.

Anyone have any idea.. how to tackle this problem.

Not all parts are identified material tank conducting or not conducting. when conducting potential difference may trigger system.
I build many of these sensors base on impedance for water up to 95 deg Celsius.
It works for most of the fluids however flammable fluids not advised.
Use a 10Khz aprox signal generated by your mpu couple this via a cap into the water( rvs pole free hanging from top to bottem) and feed output back via opamp into mpu. The opamp is also protecting the mpu input.
Ground the tank to gnd mpu and connected to a second rvs pole in the tank distance 5cm approx. from the generator rvs pole.
the ph of the water is related to the gain high conductance less gain.
capacitive ac coupling does avoid the galvanic problems (eating your sensor pole).
maximum signal into opamp is tank empty hence impedance is minimum.
Accuracy level 5% approx.