Did you check those sensors to see if they have a pullup resistor?
The black lead should measure 0 volts with the sensor not activated by the water.

 

Updated schematic if sensor has pullup resistor.
View attachment 287783
Without pullup resistors. Looking at the image in post #2, don't see a pullup resistor.
View attachment 287785

OMG, it's working! (option one tested)
Sir you are genius!

 

Did you check those sensors to see if they have a pullup resistor?
The black lead should measure 0 volts with the sensor not activated by the water.

I was just about to post when I seen your post


Couple more questions
In terms of reliability is there any special parts you would use to make it very sturdy and long lasting?
I used jut parts I found around , even those diodes I have no idea what they are.
Will order the proper parts.
Is it possible to order somewhere custom PCB for this?
I will make batch of these and cast them in epoxy.

Thanks

 

Did you try option #2 just curious?

 

Did you try option #2 just curious?

You always beat me to it
Just did, also working perfect, no difference.

 

Stick with option 2 one less diode.
How many pc boards do you need?

 

Stick with option 2 one less diode.
How many pc boards do you need?

Will do
Don't know, I can't even imagine how much that costs.
I could probably use about 10 or so.
The design is so simple that I'm tempted to do the layout on breadboard.
Love it! no part is good part in terms of reliability.
Could you please comment on the components you would use?
I presume ceramic capacitor will be more reliable then electrolytic right?
I left the circuit on over night with the relay triggered and IC is not even warm
so I presume no need for transistor to drive it.

Once again thank you so much!

 

Looks like company like JLCPCB will do prototype very cheap.
Not sure how it work though if I have to draw the design or if I just provide your circuit to them,
will look in to it.

EDIT
Used EasyEDA to create the PCB, so I'm ready to order it.

 

Could you please comment on the components you would use?

Mainly where you purchase the components like DigiKey or Mouser. None of the parts are critical. Electrolytic or tantalum for the cap is fine, 1/8 watt resistors will suffice.

 

Mainly where you purchase the components like DigiKey or Mouser. None of the parts are critical. Electrolytic or tantalum for the cap is fine, 1/8 watt resistors will suffice.

That's great!
Many thanks!

 

Update, and some additional questions.
Got all the parts, ordered PCB one design with PS on board, and the other separate.
Assembled both and here are my questions:
The one with the PS on the board (HLK-5M05) is not working right.
I have to hold the upper or lower lead to the sensors with my hand
and then it works right otherwise it will only once switch the relay and that's it until I touch one of the sensor leads.
Desoldered the PS from the board powered with separate 5V PS, problem persists.


The one without the PS is powered by one of these Ali specials

Works great except I noticed that I have to keep the PS minimum 3/4" away from the PCB
otherwise the PS often resets when the relay is switching on/off power to 120V/5.5VA solenoid valve.
No big deal but strange.

Can someone please chime in on these issues?

Thanks!

 

The first issue sounds like the wires to the sensors are acting as an antenna and by holding the wires your body capacitance is reducing the interference, but doesn't make sense that the second board does not have the same issue when using the same setup. Possible the pc board layout may be the problem
The second board with the different power supply is probably getting some interference from that switching power supply since the problem is eliminated when moved 3/4 inch.
Couple of things to try are:
Replace the C1 caps with 100uf. If that doesn't help then add C2 a 10uf capacitor from pin 4 to ground.

 

Which version schematic are the two boards using? The one with 2 diodes or the other with only 1 diode?

 

The first issue sounds like the wires to the sensors are acting as an antenna and by holding the wires your body capacitance is reducing the interference, but doesn't make sense that the second board does not have the same issue when using the same setup. Possible the pc board layout may be the problem
The second board with the different power supply is probably getting some interference from that switching power supply since the problem is eliminated when moved 3/4 inch.
Couple of things to try are:
Replace the C1 caps with 100uf. If that doesn't help then add C2 a 10uf capacitor from pin 4 to ground.
View attachment 289638

Interesting, will try that today.
The second board has completely different layout so that makes sense.

Using the one with 1 diode only but tried to add the 2nd diode and it did not make any difference.

Thank you!

Update
Tried with each and both caps as you suggested.
No difference.
My touch still works in both cases

 

Post a HD close up photo of both sides of each board please.

 

Post a HD close up photo of both sides of each board please.

Pretty sure noty very efficient but I did my best

 

Thanks, but I would like to see an actual photo of both boards both sides.

 

With components on it will be a problem since I "love" my hot glue

BTW also noticed that my choice for the 5V relay (JRC-21F ) isn't great, after while switching the 120V/5.5VA it starts to shiver.
The problem is there only when the contact voltage/load is applied since without it it worked perfect for hours even days when I had it on the bench.

 

I wanted to see the actual traces and components on the boards.
For the relay a 10uf cap across the coil might help.
Another though is to place a 100nf cap from pin 5 to ground on both boards.

 

Will try that! Thanks!
Here is couple pics of my nasty workmanship
Those extra wires are there to connect reed switch to disable the circuit when tank is removed.