4 ms·

How do you keep water touching sensors working long term? I tried similar sensors but they all get rusted / oxidized to work properly after a certain time.

ljlukkar 3y ago
You need industrial level sensors and the water needs to be flowing constantly through them. I built something similar about 15 years ago and tested many sensors. In the end I had to pay about 1000 dollars for ph and ec meters that did the job reliably. To be honest there is nothing new here. This is how big greenhouses have been operating for decades. In small scale there is more work maintaining the automated setup and calibrating the sensors than it would take to do the measurements and dosing manually.
- steve_adams_86 3y ago
  I think it can cost quite a bit less now, but you’re right — it isn’t cheap.
netsectoday 3y ago
You can use capacitive water sensors taped to the outside of non-capacitive containers (aluminum foil, a resistor, an arduino, and a plastic 5 gallon container), but honestly all you need are DNI timers to "automate" any grow operation. Put your lights and pumps on a schedule and there is absolutely no reason to get more creative. If you do anything besides low-level timers you're making it complicated and brittle with no added benefit.
steve_adams_86 3y ago
These sensors are designed to withstand contact with water and to minimize hydrolysis, and I haven’t had issues with that so far. I’ve been running this system for close to a year and they still seem to calibrate just fine.
rytis 3y ago
not OP, but one of the tricks is to activate the sensors only when measuring, so there's no constant DC applied to the sensor wires/pads. once you have that, reduce measurement frequency, so to mainimise the time when voltage is applied to the sensors. for example once every hour for moisture is sufficient, and 1/sec isn't really going to help much.
- amelius 3y ago
  Would alternating the polarity work?
  - jesprenj 3y ago
    Off-topic but perhaps interesting: That's what they do when performing catheter ablation (a medical procedure for curing cardiac fibrilation by destroying minute parts of muscle with electric current). DC would work just as fine on this procedure, but due to electrolysis of water, oxygen and hydrogen bubbles would form, which could get stuck somewhere. Using a square wave AC quickly reverses the reaction every period, like you suggested for the moisture meter. https://en.wikipedia.org/wiki/Catheter_ablation#Technique https://en.wikipedia.org/wiki/Catheter_ablation#Technique I don't know the answer to your question, but it would be worth trying.
hksoftware 3y ago
I've seen a clever setup with the sensors in a dry container above the water tank. There is a hole in the bottom. Before testing, a pump fills the container up with the tank water, flooding the sensor probes. When the pump stops, the water drains back out into the tank.
- ljlukkar 3y ago
  The ph sensor will die fast if it the membrane is kept dry.
- steve_adams_86 3y ago
  You’d need to wash the sensors and return their caps with protective fluids. It would be totally possible to automate, but perhaps the same overall cost as buying industrial grade sensors which can handle long term submersion. You’d also need to ensure the caps contained enough storage solution at the right concentration. Over time the probes would introduce drops of nutrient solution (unless you rinsed them with distilled water, in which case you’d dilute the storage solution), and you’d need to replenish it.

ljlukkar 3y ago

You need industrial level sensors and the water needs to be flowing constantly through them. I built something similar about 15 years ago and tested many sensors. In the end I had to pay about 1000 dollars for ph and ec meters that did the job reliably. To be honest there is nothing new here. This is how big greenhouses have been operating for decades. In small scale there is more work maintaining the automated setup and calibrating the sensors than it would take to do the measurements and dosing manually.

steve_adams_86 3y ago
I think it can cost quite a bit less now, but you’re right — it isn’t cheap.

netsectoday 3y ago

You can use capacitive water sensors taped to the outside of non-capacitive containers (aluminum foil, a resistor, an arduino, and a plastic 5 gallon container), but honestly all you need are DNI timers to "automate" any grow operation. Put your lights and pumps on a schedule and there is absolutely no reason to get more creative. If you do anything besides low-level timers you're making it complicated and brittle with no added benefit.

steve_adams_86 3y ago

These sensors are designed to withstand contact with water and to minimize hydrolysis, and I haven’t had issues with that so far. I’ve been running this system for close to a year and they still seem to calibrate just fine.

rytis 3y ago

not OP, but one of the tricks is to activate the sensors only when measuring, so there's no constant DC applied to the sensor wires/pads. once you have that, reduce measurement frequency, so to mainimise the time when voltage is applied to the sensors. for example once every hour for moisture is sufficient, and 1/sec isn't really going to help much.

amelius 3y ago
Would alternating the polarity work?
- jesprenj 3y ago
  Off-topic but perhaps interesting: That's what they do when performing catheter ablation (a medical procedure for curing cardiac fibrilation by destroying minute parts of muscle with electric current). DC would work just as fine on this procedure, but due to electrolysis of water, oxygen and hydrogen bubbles would form, which could get stuck somewhere. Using a square wave AC quickly reverses the reaction every period, like you suggested for the moisture meter. https://en.wikipedia.org/wiki/Catheter_ablation#Technique https://en.wikipedia.org/wiki/Catheter_ablation#Technique I don't know the answer to your question, but it would be worth trying.

hksoftware 3y ago

I've seen a clever setup with the sensors in a dry container above the water tank. There is a hole in the bottom. Before testing, a pump fills the container up with the tank water, flooding the sensor probes. When the pump stops, the water drains back out into the tank.

ljlukkar 3y ago
The ph sensor will die fast if it the membrane is kept dry.
steve_adams_86 3y ago
You’d need to wash the sensors and return their caps with protective fluids. It would be totally possible to automate, but perhaps the same overall cost as buying industrial grade sensors which can handle long term submersion. You’d also need to ensure the caps contained enough storage solution at the right concentration. Over time the probes would introduce drops of nutrient solution (unless you rinsed them with distilled water, in which case you’d dilute the storage solution), and you’d need to replenish it.