Hi all. I'm new to this site, my English is not marvelous, so go easy on me . Hope you all understand me.

So here is my problem. I want to create low power water level sensor. Was searching for that kind of sollution on internet but didn't find any.
There are a lot of water level sensors, but most of them are high voltage or high current sollutions.
There is also a huge selection of low power microcontrollers and batteries, so this is not an issue.

Project requirements:
- non-rechargeable 3,6 V battery power supply, was looking for LS 14500
- MCU for data processing, most time it will be in sleep mode, will wake up once every two days or even less
- also have sensor dimensions limitations, height about 10 cm and max. 3 cm width
- lifespan at least 1 year

Sensor will be put in 1 liter tank. Accuracy is not that relevant, i mostly want to know when the tank is empty.

Would love to hear some suggestions and ideas about measuring water level.

 

If it is simply a certain level such as 'Tank Empty' then a simple float switch should suffice, especially for a 1 litre tank capacity.
It is also very simple to make a sealed one if necessary.
Depending on what is required to control, it could be mostly hard wired discrete devices.
Max.

 

Yes I was thinking abount that too, but I also want between measuments of water level. Well I could also have multiple float switches but it would very inaccurate (if I had 10 pins I would have 10% accuracy, but I want higher resolution).
Could I measue capacitance between two electrodes? I know it is very temperature sensitive, so i would have to measure temperature as well.

 

Have you considered an ultrasonic range finder solution. Mount the sensor above the tank and periodically measure the distance between the sensor and the top of the water.

 

I have, but they have high input voltage which I do not have or they have high current consumption which I cannot afford. :/ At least I did not find the one it is suitable for my project.

 

Yes I was thinking abount that too, but I also want between measuments of water level. I know it is very temperature sensitive, so i would have to measure temperature as well.

Well that sort of contradicts this " i mostly want to know when the tank is empty".
Max.

 

Indeed . I wanted to say that most important task is to know when the tank is empty, but I also want to know what is going on when the tank is not empty, if you know what am I saying. Sorry for my bad explanation.

 

What do you consider high current? 15mA?
Is 5V too high a voltage?
Have you seen this?

 

A simple reed switch float should work. You can make or buy them with 2 or more levels.

http://www.chicagosensor.com/HowFloatSwitchesWork.html

 

have, but they have high input voltage which I do not have or they have high current consumption which I cannot afford.

If you are only doing the measurement infrequently, like every couple of days, then you only turn on the power to the sensor when you are doing the measurement, thus even if it draws relatively high current for that short measurement period, the drain on the battery would be small.
The power can be readily switched using a transistor controlled by the microprocessor.

 

Yes I was thinking abount that too, but I also want between measuments of water level. Well I could also have multiple float switches but it would very inaccurate (if I had 10 pins I would have 10% accuracy, but I want higher resolution)........

So, how do you want to output the data? Do you want to log it over time or just the current level? How do you want to be warned when the tank is "empty"?

Ken

 

What do you consider high current? 15mA?
Is 5V too high a voltage?
Have you seen this?

Yes have seen it, but I was planing on going with 3,3 V power supply (battery). Well I could go for 5 V battery but, then I will need buck converter which will drain my battery even more, even though it has over 90% efficiency. You think the 5 V, 1,5 Ah battery will last at least 1 year if the MCU and sensor are active once every two days? This might work right?

If you are only doing the measurement infrequently, like every couple of days, then you only turn on the power to the sensor when you are doing the measurement, thus even if it draws relatively high current for that short measurement period, the drain on the battery would be small.
The power can be readily switched using a transistor controlled by the microprocessor.

Oh ok. So I have same question for you. You think the 3,6 V or 5 V, 1,5 Ah battery will last 1 year if the MCU and sensor are active once every two days? And the MCU is in sleep mode (draining uA + sending data over bluetooth some mA) most of the time. But I don't have the sensor yet, that is the problem. I don't know the drain current yet.

So, how do you want to output the data? Do you want to log it over time or just the current level? How do you want to be warned when the tank is "empty"?

Ken

Yes I did not mentioned that. I will send data through bluetooth. Every time it measures level, MCU will send data over bluetooth. When the tank will be empty it will just do the same thing as before, measure water level once every two days. I will just send data string that the tank is empty.

A simple reed switch float should work. You can make or buy them with 2 or more levels.

http://www.chicagosensor.com/HowFloatSwitchesWork.html

Have seen this before yes. But I was only able to find high voltage sollutions. Maybe if you have some specific information (operating voltage and current) or actual product link. I'm interested in 3,6 V sollutions (low power).

Thank you all for such quick response. Really appreciate that. I'm still open to suggestions and ideas.

 

If you replace the pointer with a potentiometer at its shaft, it will give you all you want at the voltage of choice, as will work as a voltage divider :

----> https://www.amazon.com/fuel-gauge-John-Deere-AM35120/dp/B000NE4SSI

1M ohm or more linear potentiometer will drain the least power for long battery life

 

Why not one of these: https://www.parallax.com/product/29131
Resistance based (150Ω/inch) linear output, so drive voltage is not an issue. They come in a couple of different lengths. It would only be powered momentarily for a reading.

Ken

 

If you replace the pointer with a potentiometer at its shaft, it will give you all you want at the voltage of choice, as will work as a voltage divider :

----> https://www.amazon.com/fuel-gauge-John-Deere-AM35120/dp/B000NE4SSI

1M ohm or more linear potentiometer will drain the least power for long battery life

Hmm yes, this is not a bad idea. Could actually build one of those myself. Thank you.

Why not one of these: https://www.parallax.com/product/29131
Resistance based (150Ω/inch) linear output, so drive voltage is not an issue. They come in a couple of different lengths. It would only be powered momentarily for a reading.

Ken

This is quite impressive. The price is pretty wild. It gave me a pretty good idea. Haven't seen that before. Will research this a bit more. Thank you.

 

You think the 3,6 V or 5 V, 1,5 Ah battery will last 1 year if the MCU and sensor are active once every two days? And the MCU is in sleep mode (draining uA + sending data over bluetooth some mA) most of the time. But I don't have the sensor yet, that is the problem. I don't know the drain current yet.

Once you know the total current drain and the time it takes to do the measurement, the total current drain over a year can be readily calculated.
Otherwise it's just a WAG.

 

Define your tank size and the material in the tank.

One simple solution or small tanks, is reed switches or hall effect switches with a magnetic float. Pressure is another way of measuring height.
Ultrasonic is usually used for large tanks. There are capacitive sensors that can be used too. The chosen technologies depend on tank size and what's in it.

 

Once you know the total current drain and the time it takes to do the measurement, the total current drain over a year can be readily calculated.
Otherwise it's just a WAG.

Let's estimate a bit. If I had total of 30 mA drain current for about 5 seconds (active state) once every two days, and the rest of the time will drain 300 uA with 3,6 V 1,2 Ah battery, for how long could I do that? Can I neglect those 30 mA?

Define your tank size and the material in the tank.

One simple solution or small tanks, is reed switches or hall effect switches with a magnetic float. Pressure is another way of measuring height.
Ultrasonic is usually used for large tanks. There are capacitive sensors that can be used too. The chosen technologies depend on tank size and what's in it.

It's a tank sized about 15 cm x 10 cm x 10 cm (length, width, height), about 1 liter of water will be put in tank. Tank is made of plastic. On top of the tank will be a hole with diameter around 3 cm.
Will look up reed and hall effect switches. Thank you sir.

 

Let's estimate a bit. If I had total of 30 mA drain current for about 5 seconds (active state) once every two days, and the rest of the time will drain 300 uA with 3,6 V 1,2 Ah battery, for how long could I do that? Can I neglect those 30 mA?

It's a straightforward calculation.
The total ampere-hour drain in a year would be (5s / 3600s * 30mA * 365/2) + (0.3ma * 365*24) = .0076 + 2.63 = 2.64 Ah.
So yes, you can neglect the 30mA if it's on for only 5 seconds every two days.
Your 1.2 Ah battery would thus last about 1.2/2.64 * 12 = 5.4 months.
So to go at least a year you will need a larger battery or reduce the 300μA quiescent current.

 

It's a straightforward calculation.
The total ampere-hour drain in a year would be (5s / 3600s * 30mA * 365/2) + (0.3ma * 365*24) = .0076 + 2.63 = 2.64 Ah.
So yes, you can neglect the 30mA if it's on for only 5 seconds every two days.
Your 1.2 Ah battery would thus last about 1.2/2.64 * 12 = 5.4 months.
So to go at least a year you will need a larger battery or reduce the 300μA quiescent current.

Thank you sir for explanation.