Good day everyone.
I'm here to ask about any suggestions, ideas and professional guidance regarding my project about vertical temperature stratification using arduino in a lake.
My problem is, I dont have ideas on what is the best sensor that I can use to determine the temperature under water, with different depth, that will last for about 3 to 5 months data logging.
Can anyone could suggest to me what sensor should i use?
First consideration : DS18B20
There is an encapsulated version, which is the obvious one to buy, but it is waterproof only up to a point. You will need to get liberal with epoxy or something.
The advantage of the DS18B20 is that you can put many sensors on one Arduino pin.
If you can get them similarly packaged, you may find thermistors are entirely adequate.
The depth of the lake may have some bearing on your choice.
This does not appear to be reasonable to me. 20C that is about 1.6 degrees C (+-0.8). I would expect the gradients to be in the 0.01 degree range but I know very little about what you are measuring.
This will be in a building?
That is ~5K byte. FRAM may be your best bet as it is low power, non volatile, and fast.
Later,
beware
Those you found maybe ( they look good ) but many sold as waterproof are not , despite what it says , the tip is , but you cannot submerge the cable - a few you can , but many from Amazon etc you cannot .
Pressure rises rapidly and they may not work at depth ??
You can make a thermowell from a length of copper tube , seal the end with a solder fitting and fit your sensors daisy chain and lower them inside the tube .You can back fill with a silicon oil if you wish .
Or ( better still ) make up for a single sensor with a capped short length of copper tube joined to a plastic length ( use plumbing fittings for this ) .
Submerge !!
If OP never worked with DS18B20 before, each one come with a unique serial nr. Connect one at a time, take note of the s/n and label the sensor so you know the correlation. Otherwise you'll end up with a bunch of sensors and not knowing at which depth each and one are at.
Supplemental question.
If your mounting arrangement were a single vertical 'stick' to be lowered into the water, how many sensors would be mounted on the stick, and what is the range of positions? Your answer might be something like:
21 sensors, 0.5m spacing, 0 - 10m range.
People seem to imagine this as something to be lowered into water. I would go the opposite way and build all into a sealed, water proof box and anchor to the bottom. The temperature sensors would be floated in line to the surface. I would have a multi-conductor cable to the shore that provides power and receives data from the box on the bottom of the lake.
Correct, of course! But if the OP is looking for an autonomous monitoring device, he had better be prepared to build several so at least one will work for the required time. By having a shore connection the underwater device can be monitored real-time.
I either never found a thermowell, or never found a thermowell I was prepared to pay for, so I made my own. I'm not sure a thermowell in the accepted sense is what you need anyway. A thermowell is inserted into a wet tank or pipe, and the sensor is kept dry on the outside. In your case I think you intend the sensor dry inside a pipe which is wet on the outside - something totally different. I imagine you need no more than a short length of pipe for each sensor with a compression gland at each end. It would also act as a sinker.