Swimming pool level detection

A french designed swimming pool level detection that I have in use employs a sensor with 3 stainless steel electrodes. Because these electrodes started corroding after a few years I wanted to find out how this system operates. Through reverse engineering I managed to draw the circuit attached.

My question is to know what your comments are about this design in relation to improving the concept, or maybe to get altogether different concept ideas from you.

After studying this for a while I think to be able to explain its workings as follows:
1. red-red and blue-blue are 220V ac inputs
2. red, blue and yellow are sensor wires (basically 3 stainless steel bolts spaced vertically about 2cm apart each (I think red first -lowest-, then blue, then yellow).
3. black and blue-blue are shorted when the electrovalve needs activation: blue-blue and red-red go to a 220V-24V transformer for the valve
4. when the sensor resistance drops more voltage comes across the secondary 12V transformer, hence -Vcc increases to a higher negative voltage
5. Q2 starts conducting, collector Q3 goes high, Q6 starts conducting and K1 switches from blue-red to blue-yellow
6. if Q2 starts conducting the 4060 gets Vcc through Q2 and starts to "oscillate" (I think at a very slow rate: K2 must be switched on and off very slowly
7. when 4060 output Q5 is high the electrovalve is activated.

If I made any errors in my reasoning above please let me know?

I would very much appreciate your inputs, either to improve this design, or to suggest new concepts. And yes I know: corrosion is due to the electrical current through the stainless steel sensor bolts.
Thanks!

unrelated to your circuit
I once stuffed a 1/2in pvc pipe full of tinfoil with a wire leading to the pin, then loosely wrapped a single turn of grounding wire around the bottom of the pipe at the endcap (just to hold the wire in place under water.. ) creating a capacitor whos capacitance changed with its depth in water
with 5v adc on a microcontroller i was able to read the response accurately enough for my liking , however i have no comparison to know how it actually holds up

I

Dr_Doggy said:

unrelated to your circuit
I once stuffed a 1/2in pvc pipe full of tinfoil with a wire leading to the pin, then loosely wrapped a single turn of grounding wire around the bottom of the pipe at the endcap (just to hold the wire in place under water.. ) creating a capacitor whos capacitance changed with its depth in water
with 5v adc on a microcontroller i was able to read the response accurately enough for my liking , however i have no comparison to know how it actually holds up

Click to expand...

If I understand you well, you had the foil in the pipe isolated from the water and the piece of wire exposed to the water? At what frequency and voltage did you run the sensor?

In my case I already have a sensor with 3 probes, although that gives me a digital result rather than an analog as in your case, I would like to stick to that sensor because it is good enough for me. But what I prefer would be a microcontroller signal driven to these sensor pins, ie a square wave that is either shorted (water level = ok) or open, and then only at intervals instead of a continous sine wave as is the case now (resulting in corrosion after a few years).

What I also do not like about this schematic is that the 220V ac is only a transformer coil insultation away from my pool. I am surprised this is even considered by the designer?

Morning .. .. ..

Can I ask .. what are the 'extremes of your measurement ? If you're looking to detect a drop in water level, what is the distance between full and the lower level ?

The reason I ask ?

earckens said:

I would very much appreciate your inputs, either to improve this design, or to suggest new concepts.

Click to expand...

A very different concept .. .. .. a capacitive probe .. ..

This is a small prototype I built to test the theory, but I did eventually build a 40 cm long probe which currently displays the water level of a 50 litre water butt on a analog moving coil meter.

The probe is made up of an enamelled wire loop supported by a threaded brass rod. The enamelled wire and the brass rod produce a variation in capacitance according to the water level which, with the aid of a voltage follower is displayed on the moving coil meter.
The main advantage is that because there is no direct contact with any voltage and the water there is no electrolysis.
Although it looks fairly primitive, as anything I build would have to be, it is surprisingly stable and accurate.



This is the circuit board.
The scale of measurement is varied by the combination of capacitors viewed in the top left hand corner of the PCB.
I have to say, the design is not mine .. .. .. I do have a schematic, but I will seek the permission of the author before publishing it.

Just a thought !

S

I am curious to see the schematic.
What output signal do you get (min, max, dc?) for given water levels (empty, full, in between)?

Hi

I've just discovered that the author published his schematic here .. .. ..


https://www.electro-tech-online.com...ill-run-to-an-led-display.107910/#post-883946

I'm not sure I can remember exactly what the output signal was, looking at my notes, I've written down 1.1 - 1.7 vdc, but ... memory is not what it used to be.

I actually displayed the results on this ..



So it won't have been much.
Of course, the other advantage is the whole thing is driven by a PP£ (9v) battery.

S

Interesting schematic: duty cycle changes in function of capacitance.. I already have a microcontroller program that is able to measure duty cycle width (using direct register addressing, very fast), however this will require some extensive testing and prototyping. Good for next winter.
Right now I would prefer a quick solution before summer arrives, so probably still resistance driven based on my current 3-probe sensor.

Nifty little circuit, no connection to the liquid is a good way to avoid corrosion.

If you want to minimize corrosion on the electrodes avoid dc voltages, preferably using a cap in series with them, a long time ago I copied someones circuit for a level probe, it used a 4052 analogue multiplexer to switch the probes around (I'd do it different now).
Another way that seems to have grown popularity in industry is what manufacturers like to call guided wave radar, actually just a time delay reflection measuring circuit, you transmit a pulse down a wire and wait for a reflection, the time taken to get the reflection is 1/2 the distance from the sensor circuit to the material being measured.

dr pepper said:

(...)
If you want to minimize corrosion on the electrodes avoid dc voltages, preferably using a cap in series with them, a long time ago I copied someones circuit for a level probe, it used a 4052 analogue multiplexer to switch the probes around (I'd do it different now).

Click to expand...

Even AC corrodes: mine were rust after 2 to 3 years.

I am considering a setup with measurements just once an hour or so; this pool will not drop faster and this will be a lot better than continously sending current through the probes.

dr pepper said:

Another way that seems to have grown popularity in industry is what manufacturers like to call guided wave radar, actually just a time delay reflection measuring circuit, you transmit a pulse down a wire and wait for a reflection, the time taken to get the reflection is 1/2 the distance from the sensor circuit to the material being measured.

Click to expand...

Yes, and then try to make this "kids-in-the-water-splashing-fools"-proof..
Too complicated... just resistive or capacitive measurement I start to think..

I use the capacitive probe to monitor the water storage level at my caravan - our old caravan, if the pump ran dry it would run all day without a problem. However, our new caravan if pump runs dry, you have about 50 secs before it burns out !

I've also had some success monitoring the fabric conditioner tank levels with an ultrasonic transmitter / receiver but moisture is a problem.

I'm sure you're microcontroller program is much more sophisticated, I just thought maybe simple is best ... .. and I'm certainly that !

S

I've seen commercial capacitive sensor water level measuring units on pleasure craft for the same thing, displaying the contents of the water tanks.
I'll add that schem to my ideas folder, using a 7555 and hc logic would probably make the thing run on next to no power and be on constantly while the van/boat is in use.

I thought about doing this a few years ago. I was going to use a capacitive sensor and use the comparator of a pic as an oscillator to measure the capacitance. Would make an interesting project.

Mike.

Pommie said:

I was going to use a capacitive sensor and use the comparator of a pic as an oscillator to measure the capacitance. Would make an interesting project.

Click to expand...

Sounds much too technical for my simple mind ! Nice thought though .. .. .. .

S

There are some examples where folk have done that with a pic to make a capacitance meter, some alteration work would end up with a useable level meter.

This is a basic over view of several methods of liquid level measurement including the bubble method which is one I like to use. Using a pressure sensor and a small air pump, in many cases a small aquarium pump will suffice, we can get a good representation of water level in a tank. Here are several good examples of using a bubbler to measure a tank or swimming pool water level. In the case of a closed tank, for example a closed tank under pressure I like using a differential pressure gauge type system. Board mounted simple basic pressure transducers (sensors) like these from Honeywell and similar are relatively inexpensive and serve good enough for most applications. One merit of using sensors like pressure and differential pressure is getting away from using alloys which become corroded over time as a result of electrolysis and I have seen even electrodes using low AC currents become corroded.

Ron

As this subject surfaces often, I had saved a site promoting a different method to measure levels.
Check its goodness or not :
----> https://www.vegetronix.com/Products/AquaPlumb/
It adds to general knowledge of what is available on the market.

Externet said:

As this subject surfaces often, I had saved a site promoting a different method to measure levels.
Check its goodness or not :
----> https://www.vegetronix.com/Products/AquaPlumb/
It adds to general knowledge of what is available on the market.

Click to expand...

Here is another one, using a pressure sensing element: https://milonetech.com/t/standard-etape-liquid-level-sensor

I guess for the time being I will be using my actual sensor, put a microcontrolled square wave to it and do some programming with the necessary delays to compensate for waves, lag and lead times (a pool that is draining or being refilled goes rather slow). Next winter will be JimB's capacitive schematic testingtime.

Where I work there is a oil tank for heating the plant, its contents guage uses pressure, it has a mini compressor inside it, not quite sure how it works, maybe it just records the max pressure, the sense tube goes into the bottom of the tank.

dr pepper said:

Where I work there is a oil tank for heating the plant, its contents guage uses pressure, it has a mini compressor inside it, not quite sure how it works, maybe it just records the max pressure, the sense tube goes into the bottom of the tank.

Click to expand...

If I pressurize the tank and measure the difference (delta) pressure between the bottom under the weight of the fuel oil and the top above the level of the fuel oil the delta pressure will be representative of the level of fuel oil in the tank. If the entire tank is pressurized it won't matter as the delta pressure will represent the level in the tank. If the tank was water it would be inches of water and 1 PSI would be 27.67 or 27.7 inches of water. Off hand I don't know what 1 PSI of oil column of fuel oil would be but it would depend on the type fuel oil. As long as I know the delta pressure I can derive the tank level.

Ron

Makes sense, would temp affect the oil's mass?