Hi guys!

I think this is my first post and I'm a beginner regarding electronics, in general. So I want to build a water sensor for my fish tank, the water evaporates rather quickly and I need to add water twice per week. The tank is small (30 liters), and the amount of water need to add usually is 1,5/2 liters.

So in order to automate this procedure, I've tried the circuit in the image below, using a 555 timer as a comparator and 2 BC547. Instead of using a LED to indicate the water level is low, I'm going to use a 12V DC water pump.



My main question is, since I'm using a probe made of 3 wires, is it safe to put in the water with livestock? I will protect the probes so fish wouldn't to touch it, but even though I'm not sure how safe it will be. Could you please help me out? I've tried the circuit in a glass of water and works as intended.

Do I need to limit the current from the circuit or the current is so low that doesn't make a difference?

Best regards,
João Dias

 

I don't know whether 12V is safe for fish but maybe you could shield them from the voltage by making the VCC 'probe' a metal mesh basket which encloses the other two probes. If your circuit has no other connection to ground then the fish would see no voltage.

 

Hi Albert!

Thank for you help. I've made a drawing to try visualize your suggestion, but it seems almost the same I was planning to do with a PVC pipe, in order to keep fish away from the probe.



Did I drew right or did I misinterpreted?

I think I need to provide at least 8V or 9V because of the minimum requirement for the BC547.

 

What is the requirement of BC547 that needs 8V?
One should be able to have a circuit working even below 5V.

 

What is the requirement of BC547 that needs 8V?
One should be able to have a circuit working even below 5V.

As I've seen in the datasheet the minimum Emitter – Base Breakdown Voltage was 6V, but in other manufacturers saw 8V

The probes themselves may be toxic to the fish depending on the alloys used could wipe out your tank

Well that's even more important that I was thinking, copper is very dangerous and I didn't take that in consideration. Thank you so much for the tip, that is very important. I'll try to see if there is any conductor better suited for this application. If I don't see a solution, I'll try using a way to find a way to see the water level in a "mechanical" way, through switches or magnets.

 

The problem with the scheme is the inevitable erosion of the probes. Electrolysis of water requires only 1.23V (in the minimum case) and the probes will be adding metal ions to the water at some rate. It is unavoidable.

You should use a float switch which is isolated, or possibly something like this capacitive sensor which is fitted to the outside of the tank entirely.

 

The probes themselves may be toxic to the fish depending on the alloys used could wipe out your tank

Yes, good point.
Maybe the probes could be carbon rods?

 

You are misinterpreting that data. It means that exceed 6V and you will break the transistor.

As for electrode material, you need to use carbon or platinum.

 

You are misinterpreting that data. It means that exceed 6V and you will break the transistor.

As for electrode material, you need to use carbon or platinum.

In an automated electrolysis of water demo I built I had to use platinum electrodes. I was glad I wasn't paying for them.

 

I think, if you use lead for the probes, electrolysis would result in some lead being dissolved in the water and I doubt that would be good for fish. That's why we now have to use lead free solder to stop it getting into rivers.

 

Lead is commonly used in aquariums to hold down plants -done it myself for many years, just remember to remove and replace some of the water from the bottom of the tank every week or three, which I am sure you already do already.

You need to put 100k to 1 Meg of resistance between pin 2 and ground (pin 4) so that your trigger input goes to ground and is not left floating when the water gets too low.

The capacitive sensor to which Yakov linked in post #7 is an excellent idea -no currents, no corrosion possibilities, no worries.

 

Lead is commonly used in aquariums to hold down plants -done it myself for many years, just remember to remove and replace some of the water from the bottom of the tank every week or three, which I am sure you already do already.

But in that case there is no electrolysis current to dissolve the lead into the water.

 

Thank you so much for all the information! I've never thought so much people would reply! I will check the capacitive sensor Yakov suggested, it is probably the best way, no corrosion, lead poisoning and other problems.

Regarding the suggestion for getting a Oscar fish, I would be so sad seeing all my fish and shrimp being devoured by that little monster!

 

Thank you so much for all the information! I've never thought so much people would reply! I will check the capacitive sensor Yakov suggested, it is probably the best way, no corrosion, lead poisoning and other problems.

Regarding the suggestion for getting a Oscar fish, I would be so sad seeing all my fish and shrimp being devoured by that little monster!

Capacitive sensors are really the best when you have a non-metallic container. The ones I linked to are just a random choice, you can find them in all sorts of configurations from many sources. I was just looking for a smallish one, there may be smaller ones available.

Good luck.

 

Capacitive sensors are really the best when you have a non-metallic container. The ones I linked to are just a random choice, you can find them in all sorts of configurations from many sources. I was just looking for a smallish one, there may be smaller ones available.

Good luck.

Yup, I've ordered 2 of these, it is the same capacitive sensor and seems to work just fine. Thank you for the tip! I'll hook it up with a ESP32 or 8266 to get real time data of the tank and control the refilling process as well as temperature, etc..

 

In the interest of getting away from exposed probes I would think about a simple float sensor using a few small magnetic reed switches and a floating magnetic ring. A Google of magnetic float switch should get you started.

Ron

 

I'm late to the party, so this might already have been covered.

The circuit in #1 will not work. You need a pull-down resistor on the Trigger input. also, the Threshold input is permanently grounded. You need a resistor from the emitter to GND.

Currents through the water are very low. Best to use a CMOS 555 variant. LMC555, etc.

ak

 

The TS said it worked testing in a glass of water. Nothing wrong with the threshold pin tied to ground that's what the reset pin is used for. Resistor on pin 2, not needed.

 

What pulls the Trigger input low when the probe is exposed? The Trigger input is an unbiased PNP transistor base. The datasheet says it needs a current (out, not in) of 500 nA typ. Other than leakage current inside the chip, I don't see a path for this in the circuit. That is an equivalent resistance of 6 M to GND. I suppose it could be a leakage path through the circuit construction or something in the test setup wiring. But for far more reliable operation, it is better to provide an explicit current path to GND for the Trigger input.

ak

 

For anybody who missed it the thread starter has decided to go with a capacitive sensor.