Thanks guys.
I am changing to a pulsed circuit (1 reading every 10 minutes) feeding latching transistors.
I really appreciate the help.

 

Certainly the function can be achieved with just a few CMOS logic gates. No programming and no processor, and not much power wasted.
AND, beware that aluminum CERTAINLY will corrode in a nasty manner. Use Stainless steel.

 

Thanks Bill. Stainless it is.

 

What controls the pulse every 10 minutes and how is it latching?

 

OK, you can do it with one IC and three transistors. A CD40147 priority encoder using the inputs one, two, and four, and the first three binary outputs. It will still require a positive 5 volts into the tank, and then the outputs each driving the base of an MPSA13 darlington transistor. Depending on the supply voltage, probably each LED will need a dropping resistor. The IC will want a five volt supply, probably. So only 7 parts unless the LEDs need the resistors. The CD40174 was in my RCA digital IC book. Not sure if it is still available. It is a decimal to binary priority decoder.
You can do the same thing with NAND gates. Actually, probably with AND gates as well. I need to stop and draw the truth table for that.

 

Thanks for the effort Bill, but it's easier with discrete components.
Regards.

 

What controls the pulse every 10 minutes and how is it latching?

My controller is:
https://www.robotshop.com/products/...3ODA1NDIkbzckZzEkdDE3NTk3ODE5NDMkajUzJGwwJGgw

 

Seems like this controller could replace the transistor circuit as well.

 

Yes, it probably could, but anyway it's max'ed-out.

 

OK but do you have or need a latching circuit for the LED indicators if pulsed every 10 minutes.

 

Not sure why individual components are easier than ICs. And certainly the ICs will consume less power. and no leads to trim. Oh Well.

 

Not for the LEDs at the tank (as per the schematic), but for remote display.

 

OK, so the LEDs are located remote and will need a latch?

 

Not sure why individual components are easier than ICs. And certainly the ICs will consume less power. and no leads to trim. Oh Well.

I find it easier when building on prototyping boards to use old school thru-hole parts.

 

OK, so the LEDs are located remote and will need a latch?

There will be a remote physical panel and a web page.

 

There will be a remote physical panel and a web page.

Do you have a circuit design for this remote physical panel?

 

Do you have a circuit design for this remote physical panel?

Not yet.

 

There will be a remote physical panel and a web page.

That's a very different situation than what you describe in your first post.

You don't need your sensing circuit to turn on LEDs at all and it doesn't matter whether more than one is active at the same time. All your sensing circuit has to be able to do is inform your controller which sensors are currently covered. Your remote display then uses this information (which can be processed in between) to turn on the appropriate LED.

 

If it helps below is a latching circuit using transistors based on the design in post #14.
Component values dependent on Vcc supply and LED current.

 

Certainly the digital logic ICs are available as thru-hole devices. Those SMT ICs are way too small to experiment with.
Iwas only suggesting the full-size thru hole ICs.