Hello

My dad asked me to make an automatic water pump activation device for our home pool. The main pool pump motor is housed in a 1,2 meter deep 'well'
and that well sometimes fills up after heavy rains so it's necessary to pump the water out before it reaches a certain height and starts damaging the main pool motor. I've got a pump which is made for this purpose but the issue is that it uses a float to detect the water level and does not trigger until the height is over 40cm which is way too much (I've already tried to think of ways how to modify it but it just doesn't work out). After consulting with a colleague we have come to a conclusion that the best option would be to use a triac and switch the pump from mains. This way no surgery/modification is required to the pump.

I have drawn an illustration of the setup. I know how I would make it trigger at the maximum allowed level but I don't know how to introduce hysteresis
to the system so that it keeps pumping until it reaches the lowest level. I've never used triacs.
The other idea that I had was to use a 'linear' feedback loop with a triac where the switching angle was proportional to the height of the water
but something tells me that water pumps won't work in a linear fashion.

The requirements are:

- Max water level : ~20 cm
- Min water level: ~7 cm

The pump runs at ~300W (Runs from mains AC plug)
I've got a triac that handles 800V 8A cont , ~60APeak

We've already come up with a semi solution but it's way too complicated. I know that it can somehow be cleverly done using
one triac and one or two transistors so if anyone has an idea let me know please. Thanks in advance.!

 

Won't you need a third wire so you can tell whether none, one, or both sensors is under water? If the third connection is via ground, and it is a reliable ground connection, and that ground is connected to your circuit ground it can help improve safety of the installation as well as enabling independent sensing of the two electrodes.

You will need to have some isolation between the sensors and the Triacs, otherwise your sensors could be a shock hazard. To get the isolation you can make a low voltage sensor circuit that is powered by a low voltage transformer and use the output of that circuit to drive an optically isolated triac driver or a relay.

Having worked with both triacs and relays, I would say that even though your application is simple, a relay would be more likely to work correctly the first time and have less of a chance of needing to be "tweaked" and would be simpler to choose and use.

The electrode-based sensor would rely on the water being conductive. Distilled water is not -might think about that while designing the sensor circuit.

Being installed near water, and particularly a swimming pool, The circuitry should be protected from splashes and weather in a weather resistant box not only to protect the circuitry, but also to protect people from coming in contact with any part of the circuit.

 

You could introduce hysteresis by using two float switches (respectively at the low and high points) operating a relay latching circuit.

 

Thanks for the answer. I do indeed rely on conductivity but I have a few more ideas that I can use. I've come up with the following solution, do you think that it will work?. I will weather proof everything.
The reason why I dismissed the idea of a relay is cost but I'm starting to look back into them. If I find a suitable one then I can just switch the motor using the relay in the circuit that I've drawn. Do you think something like this would be suitable? RELAY1 . Here is the datasheet .It costs about 2 Euro (about 3 dollars).

Sol2 (the second picture) shows the motor being switched directly by the relay

 

You could introduce hysteresis by using two float switches (respectively at the low and high points) operating a relay latching circuit.

I've dealt with that using the sequential logic circuit in the picture in the above post in block 2. What do you think? I can implement it on a NAND glue logic chip . I've got quite a few of those lying around somewhere.

Relay:
This guy seems to operate from 5V (which I will be using) . It looks suitable.

 

Won't you need a third wire so you can tell whether none, one, or both sensors is under water?

The orange wires in the diagrams are actually pairs of wires. So each sensor is an insulated pair which is exposed at the point where the measurement needs to be taken. I measured that I might be able to get an 'ON' resistance of a few hundred Kohm

 

Thank you for explaining about the wires.

Glad you are using a protection diode. Your diode needs to point the other way so that it conducts the back EMF from the coil.

The use of MOSFETs and 1 Meg resistors should make it sensitive enough if you size the electrodes correctly. It might be a good idea to add some over-voltage protection to the gates of the MOSFETs.

 

Yes , the diode seems to be facing the wrong way. I forgot about the input voltage protection. I guess A 5v zener should do the job.

Thanks for the answer

 

The relay linked to is a bit of overkill if being used just to switch a triac gate, rather than switch the pump directly. The relay coil needs 100mA!

 

Designing a triac into the circuit carries more risk of the thing not working as expected than using a relay. I think that using a relay to drive the pump directly is the most likely road to quick success.

For the case of using logic to control a triac this optically isolated zero crossing traic driver looks like a good, safe, and low cost solution.

www.ic72.com/pdf_file/m/660886.pdf