Water Tank Water Level Indicator

[CooperMan]

Now that I have fixed my washing machine pumps running on 60Hz supply when they were designed to run on 50Hz: I

[ante]

Why not a transparent tube on the outside from top to bottom so you can se the exact level at any time?

[windoze killa]

The tank is underground. Bit hard to see a tube.

You could try using something similar to a fuel tank gauge but on a larger scale.

[Bobbywolf]

What about using a mechanical float?  Basically you use a float in the tank, and a rod slightly longer than 5m being fed into a plastic Cylinder above ground.  Add some markings to the tube, and you will know how much is in it by the location of the top of the rod.  A bit low-tech for sure, but it will work.  I'm sure you can get a series of sensors to go from the top to the bottom of the tank, and electronically measure the level, but this was the first thing to pop into my head.

[ante]

Oh, I assumed there was a walk-a-round inspection path for the tank. Is it completely buryed or is its top above ground level?
A mechanical device would be the safe way to do it.

[CooperMan]

Hi folks,

Thanks for your ideas but I think I will be going with a closed bottom end 5m plastic rod with interleaved stainless steel sections, spaced about 25cm or so apart, as the water level sensors.

The bottom of the rod will also have a stainless steel (SS) section, the closed end cap, to use as the reference for all the other sensor sections of the pipe.

I would then connect a wire to each of the SS sections and use them as sensors conducting through the water to the bottom SS closed end of the pipe. These wires would be my input to the LED indicator box.

All I will need to do then is to workout an op-amp based circuit, and this is were I need more help, to switch on LEDs as the water level changes moving past the SS sensors.

I could also use the output of the LED section to alarm when the tank it getting too full or too empty. The latter being the most important as the water pump can, if there is no water in the system, tend to runaway with itself and burn out.

So what I need is for some really bright person out there to work out the op-amp circuit and I

[CooperMan]

More snaps of the water tank house and pump stuff:

CooperMan

[CooperMan]

One last snap, one can see the water in (from the roof) pipe, on the left, used to run-off the water from the roof so we can clean most of the crap. We then close this pipe and open the valve also on the bottom lift of the picture.

CooperMan

[ante]

Hi Bob,

It looks like a very peaceful place you live in, nice climate too. I know you have more or less decided which way to go but one thing came to my mind; how about an ultrasonic ruler aimed at the water surface, it would give you the exact level? I think it might be possible to tap into such a “ruler” to get an alarm level too.

[CooperMan]

Hi Ante,

Yes it is a nice place to live apart from all the pore people here; the govenment are ****, well lets not go into that here.

This is a big island off an even bigger island called Bohol and it's part of the Philippines. Map attached...

What a good Idea; I'll look into it to see what I can find but thanks for the suggestion. Any Idea about an op-amp circuit if the ultrasonic ruler idea doesn't pan out?

Regards,

CooperMan

Bob...

[ante]

Oh the Philippines, that’s nice! 8)  Well, I don’t know if op-amps are the best choice for your plastic rod; maybe this is better for several reasons: http://www.uoguelph.ca/~antoon/circ/sensor3.htm
This circuit duplicated, one for each level you want to detect and just keep the relay for the once which needs alarm. For the rest of the levels use an LED of suitable color.
Let me know what you think !  ;)

[MP]

I would then connect a wire to each of the SS sections and use them as sensors conducting through the water to the bottom SS closed end of the pipe. These wires would be my input to the LED indicator box.

All I will need to do then is to workout an op-amp based circuit, and this is were I need more help, to switch on LEDs as the water level changes moving past the SS sensors.

Hi Bob,
Here is something similar to what you requested. You could also use a LM3914 or comparators to do this.
Ante's suggestion is also a good one if you want an alarm instead of bargraph.

MP

[ante]

Hi MP,

That’s a simple circuit for the job, but will the BC108 saturate with clean water? Maybe it does, it could be worth a try.

[MP]

Yes, certainly a low cost circuit to try. I think it would light the LEDs ok.

MP

[CooperMan]

Hi folks, I have been hunting around and have found the LM3915 dot/bar display driver chip. It looks like it should be ideal for my water level detector. I have a plastic ¾ inch water pipe that I can cut into sections and intersperse stainless steel sections with a stainless steel sealed of end being the bottom of the tank water level sensor.

I see that I would need some type of resistive sensor set up between the SS sections; any ideas what this restive network might look like so I could feed the LM3915 chip to indicate the water levels.

I really would sooner not re-invent the wheel if one of you good folks has already worked on something like this. I envisage some sort of resistive chain connected to the SS sections of the pipe, but can

[MP]

Hi Bob,
Wouldn't the LM3914 be better for your project? LM3915 is a log output normally used for sound VU and LM3914 is linear.
Either way, these chips take care of the current limiting inside, so you do not need the resistors that would normally limit current to the LEDs.
Will your sensor provide different points of ground when the water level reaches the different levels? If so, one problem you will face is that the LM391x output is a low, so you will have to invert the signal.

MP

[CooperMan]

Hi MP,

I guess I didn

[MP]

I didn't fully understand your earlier post. Please disregard the comment about the LM3914 output being low since it does not matter for the water level part.
Silicon chip magazine had an article about doing this same thing using LM3914 in April 2002. Only exception: I would use 10 LEDs instead of 5 like the article.

It is here:
http://www.siliconchip.com.au/cms/A_30607/article.html
In case it does not come up properly, I have attached a diagram and copy of the article.
Hope this gets you there.

MP

LM3914_Water_Level.pdf

[CooperMan]

Hi MP,

This is just what I was looking for, thanks a million.

Just one (or two) question for now as I set about making it:

Q: Do you think that they have got the resistors, R1 through R5 in the wrong order? It looks like R5 should take the place of R2 and R2 through R4 should all move down one.

Q: What values would you recommend, for the sensor resistors, if use 10 sensors instead of the five as in this schematic as my tank is 5 metres deep?

Best regards and thanks again,

Bob… 

[MP]

Hi Bob,
Glad I could help.
I am also not too sure about the resistor ladder they have in this article. The connections to ground through the water are building a voltage divider with the different resistors. All of your sensors below the water level will be connected to ground, so you have some resistors in parallel. This gives you different resistances (calculate with the parallel resistor formula to see the different voltage levels on the voltage divider). The fourth level (sensor 5) seems to be out of synch with the others in this ladder. You might want to test some different values of resistors on the bench when expanding this to 10 LEDs. I am not sure what values I would use off the top of my head, but keep in mind that you are looking for equal voltage steps at the voltage divider on the front end (after the transistor). I would build something like this up on the bench and test resistors on a working model to see how it looks.

MP