Small Scale $10 automated watering

[Ca++]

Anyone tried automating the mixing of nutrients through perilstatic pumps?

LEDGardener had a good series going on, but he stopped.

Please share any resources

I run a RTW drip system on timers. It drips, then fills the tank again. Water usage volume is totally dependant on drip pump run time. The tank fills to the exact same level. It's like mixing the same exact thing, every single day. You will use your meters to check the tank, but you knew exactly how much feed and acid was needed, before you checked.

With this in mind, you can run the dosing pumps based on intervals. There is no need to have kit measure then dose, as you know the dose anyway. This means just dosing pumps and timers.

The biggest gain in automation, is fertigating automatically.
The second biggest gain, is having the tank fill with water again, automatically.
Chucking the food in yourself, allows for reactive adjustments. However, if you won't be there sometimes, then dosing pumps on timers can be an easy fix. This isn't easily compatible with fertigate on demand systems, though fill and dose at a fixed low level is workable. If your plants can randomly loose their watering system for an hour, while it replenishes itself.

You can use stepper motors for dosing, giving great accuracy. It seems a good idea for concentrates like the acid. However, it's much easier to buy the same size pump for everything. A feed pump will be too big for your acid, but water your acid down, an it's not. Do a little math, and see how big a tank your 5L of feed make. Then put the amount of acid needed, into another 5L container. Hey Presto, you need the same pump for each bottle, and could even share a timer, in a perfect world.
Most pumps come with hose that will melt. You need to do compatibility checks. The pipe will fail one day, regardless. So be ready for drips, and have a maintenance program in place. 'pharmed' tube seems to have the widest range of use, but see what sizes you can actually get, before buying the pumps. There are many more pumps available, than pipes.


One way to get your foot in the door, is a Aspen Standard. It's mains powered, so straight into a seconds timer. It does a nice 100ml a minute, which is just right for many applications. It comes with pharmed tube, as it's for condense removal in fridges, which could really have anything leaking in them. The tube is a phone call away. Installers often sell them for $30 which is an absolute steal. The replacement pipe costs more. It's a tough pipe to, as the unit can pump water up 10 meters of hose. Even the occasional garden pea.
If you use a single part feed, and RO, you might need nothing more.


If you have a specific question, I may have the answer. Though testing the tank, then dosing, is rather messy with cheap kit.

 

[blondie]

I had cause to pull my blueberry tonight due to many nanners. The wicks are a bit dirty but seem ok. Yes I do use just rainwater. it does seem clean water would increase life span. Here is what it looks like. Also I made a typo. Not March. Been using since May. Sorry about that. Good luck on your system. Happy growing.

 

[Ca++]

That's a good view of the base. It looks like you could double up on wicks, if you needed to. It's really quite amazing that just 4 risers can keep a pot supplied. Obviously through a bag, and the soil to must wick it up.

I feel it's all left to the gods. Who must like you more than they like me. I would come home to plants drier than that teddy

 

[blondie]

Sorry forgot one thing. I think the bottom of the fabric bag constantly being moist may deter roots from growing through the bottom of the bag. Assuming roots are growing looking for water. No need to go through the bag to find moisture. There is nothing special at the bottom of the bags. Most times using the bags roots have grown through in the past before using this system. Good day to all.

 

[exploziv]

I used all of those except the pump. The sensors need calibration and they will not last more than a few months (water enters the side of the boards and they delaminate). Also the electronocs at the top need to be protected. A simple resistance measuring circuit with 2 nails is better, you just have to switch on the power just for the time of the measurement, so they corode much slower.

 

[Ca++]

Good input @exploziv Thank you.
I use water presence monitoring, and have found most metals to last maybe a day. I was shocked how quick things just went black, and would no longer see the water they were dangled in. I started with copper wire. Quickly moved to stainless, then on to 316L which rapidly blackens, but never looses sight of the fact water is present. These sensors of mine get an AC signal which should reduce electrolysis, and 316 is marine grade, so needs to stay shiny on someones boat at sea. I suspect the P might be at play, as that's what turns rust black in hours.
What are your nails made of, and is it hydro? The metal prong type sensors get very bad press in general. I have a kit here that should measure moisture, but I'm using it for just presence. Suspecting as the water rises up the tank to reach them, the bit getting wet will blacken, and the tank need to get higher up the prongs next time. I have not run it though. It's an auto-fill kit for a friend, who I decided deserved something better.

I think I like the idea of improving the weather resistance of the encapsulated one's. Perhaps seeing how much resin you can put around it, before it's so thick, it can't see anything through it. I have a sensor on route, that see's water up to 20mm away. It's just for switching, but is adjustable and at 20mm range, lives outside the container. Different project, but same tech.


It's rough, but perhaps smear one in silicone, then pop it in heat shrink. Leave enough spare heat shrink tube at the bottom (say an extra 6") to fold it back towards (and out) the surface. So below the surface, it's just tube with the sensor inside. No joins in the tube, or seals, below ground.
Surely it will see through heatshrink to some degree.
I don't think a zip-lock bag is quite the jacket it needs. too weak to push in the ground, but it could be carefully potted in.

Don't you think it worth pursuing? I notice a lack of commercial products using any approach. Just bits of tack.

 

[exploziv]

Not sure the external sensors can be calibrated to see the level of humidity present in soil, but they do detect water from 1-2 cm away, I had them in my coffee machine and humidifier before. They seem to only see if there is water on not near them, and not much more.
The nails are the anti corosion type, not sure what they plated with? Maybe Zn? I know i pulled them out of wood after years and in my country they use green timber for construction, yet the nails were still pristine. They will corode, tho, if voltage is passed through them all the time, so I just made the code put current through them just miliseconds before they are read. Worked in coco and soil/soiless. Ofc needs to be calibrated on your soil mix. If they ever fail, you can make the code alert you on that, cause the readings will be bogus and way out of range, same as with the capacitive ones.
Honestly, getting a grow or 2 on the capacitive one might be a good start, if u buy them cheap enaugh. I have seen better sensors out there, that claim they don't corode in any way, but I am not yet inclined to spend over 50 usd on them. Maybe, one day, when they get cheaper, will have my go at them. My current aproach I am trying is using math and a bit of "stupid AI" and just watering based on temperature and humidity monitoring. That way I don't need sensors in each pot and ofc I could run it cheaper for a bigger room, provided the plants are about same size. Sometimes needs adjustment (I check on the plants anyway so no problem to press on a button if i think soil is too dry) but it is better than a timer and a pump. I think sensors will get better soon enaugh, so for now I try to put my efforts more into learning how to keep plants happier rather than tinkering with the watering systems.

 

[Ca++]

The external ones use the same tech, but to solve a different problem. For soil moisture grading, it's natural to want to stick something in the soil. While for just fluid presence, It's great we can move a bit further away. Solving the problem of having to stick something in the vessel.
I'm looking at soil probes, not so much as a grading system, but just as a means to make a logic choice of dry or not. The external one's, like me, are not grading moisture, but making the logic choice of wet or not.
I'm expecting to take the internal ones, and add to them, the circuit board the externals already have. A board that reads the signal, and switches at a point we decide.
The difference will come down to sensitivity. Being in the container, you can read soil moisture to the nearest 1% with fluctuations of 1% being detectable. Being outside the container, this resolution will be reduced. It's not as reduced as wet or not, like a binary choice. There is a scale, as both distance and fluid types can be accounted for. That fluid types adjustment might be something that chucks both types off though. We are talking EC.

The truncheon caught my attention with it's graphite electrodes.


It sounds like you are looking at RH in the air to indicate if the plants are drinking. I briefly looked at that, when the Omron 61F I ordered was the wrong version, capable of switching with different air RH. I couldn't possibly process the number of data points required to figure out what could be measured directly though. People with data logging will recognise what we are talking about. The RH peak after watering, that flatlines for a while, before the plant starts to reduce it's water usage as it's running out. It's amazing how dry a container a plant can live it. We know that living is all it's doing though. With reduced water, comes reduced growth. A plant in veg will fight back when the water is available again. It's a stress/steering mechanism. However, by week 4 of bloom, it's no longer useful, and has a negative approach. It's never so clear cut as that, but I'm chatting shit for other readers here. Expressing the importance of decent moisture control through later bloom. One user recently posted cannabinoid production can take a huge hit, if there is continued water stress. Like mine.. where I really must go water them right now.

 

[Ca++]

Anyone tried automating the mixing of nutrients through perilstatic pumps?

LEDGardener had a good series going on, but he stopped.

Please share any resources

This is either very good, or very rubbish

103.44C$ 44% OFF|Yieryi WIFI Online PH EC Meter Conductivity Temp Water Quality Tester Controller for Aquarium Soilless Cultivation Hydroponics| | - AliExpress

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

I'm not sure if I should get excited or not. It claims to adjust pH and EC, and offer a phone app to see it from wherever. £60 ($80)
This times in very well with the question of dosing pumps.

Keeping a constant tank level isn't particularly hard. You can use a tiny pump to fill a cup, no matter how low the reservoir. Then the sensors go in the cup. The cup overflows back to the reservoir.
I know my dosing takes ?minutes, so my dosing kit is turned on for ?minutes +1
That +1 gives time for the dosing operation, then kills the power to that kit, incase it's broken. My fill timer is the same. Filling stops at a float switch, but the fill pump doesn't get power for much longer than that anyway. If the float fails to stop the fill, I don't get a big flood. Thankfully it's always stopped, and my tank is big enough to take a bit more water than needed. Just not every day.

I would be hitting the 'buy' button if I actually had £60


Edit: One issue I'm yet to iron out, is stirring the bottles. Some even say 'not machinable'
Most pumps are not going in there. Things like stirring plates and magnetic stirrers are on the cards.

 

[exploziv]

Sounds like if those external sensors can be calibrated to read even a crude level of moisture in soil, without soil contact, they would be the perfect choice. Nobody needs 1024 levels from dry to wet, as those soil capacitive ones can read, 4+ levels is probably enaugh to decide when they get too dry.
That controller sure sounds nice, and it might work. Easoer to keep sensors working whrn always submerged, problem is.. do they use good sensors/good calibration?

 

[Ca++]

The external sensor is in the post, so I might get some answers before fall It's a shame the adjustment screw is actually on the container side though. If it's fudging with the sensor, rather than the measuring electronics, I won't be able to relocate it. There are a few styles of sensor though.

I didn't price up new pH electrodes, as the bnc connector makes them look pretty standard stuff. The EC probe a graphite affair, but I see a third element to the EC probe. I would say temp, if temp were not a separate probe. It's a little odd.
I'm not too concerned about the box of tricks working, but the phone app could be a nightmare. I really do prefer simplicity over user manuals. I have far too many 'smart' devices, that need more interaction than necessary to function. Considering where it's coming from, I should find a youtube before considering it further.
The 1024 steps of analogue input seems illuminating. This processor doesn't need programming to do that then. It already knows it's job is A/D conversion of the input pins. Like it's a hard wired function of the device. Not something you program it to do. From a programming prospective, you have these 1024 points already aquired for you, and you decide what to do with them?

I had not realised you knew about this sort of thing. Your help is appreciated.

 

[exploziv]

In reality you would only use a few hundred steps to get full scale, and you need to calibrate each sensor individually. You would find that depending on what medium you use for growing, 400 something might be bone dry, while 600 something is totally wet. Other sensor of same kind might be 600 for dry and 800 for wet. I have never played with the calibration screw on the sensors, tho, you might be able to bring them all into same intervals, but for me it was easier to just test each one to see what dry is and what wet is and translate that to percents. When they short or read badly they will be out of calibrated range, so you would see them showing 100% all the time usually, because they would get closer to 1024 than the useful interval that was determined. Easy to set up, and if you have spares you can count on using them for about 2 grows, on average. Let's call it 6 months lifetime. Not if u don't protect the top electronics, tho, first spray of water and they are out. I ended up encasing the tops in melted plastic from a heated plastic gun. It worked. Till the water got to them entering from the sides of the sensing "pins".

 

[Ca++]

Guy tests the one without switching. Looks over the software. 3 months later he is buying one for another tank. Likes how it alerts his phone. ph rh ec temp (2 more I forget) can all text him at his preset limits.

Another vid shows it uses 4 6.8 9.2 not 4 7 10. Not a huge issue, but it's not my standard standard.


Seems like it's real

YIERYI are a professional instrument and meter factory.

Shen Zhen Yage Technology Co., Ltd registered the brand "YIERYI" in 2016, The main products are: PH meter, Tds ec meter, Do meter, Orp meter, soil meter, refractometer. Our products are widely used in: aquariums, hydroponics , aquaculture, environmental protection, swimming pools, beauty salons...

shop.yieryimeters.com

I only see spare orp probes, despite most of their kit taking replacement pH ones as a consumable. That's awkward..

 

[Ca++]

The water presence sensor arrived. In hand, I made an observation. This sensor is maybe 15mm round. The stick for measuring, about 20mm X 75mm. This presence sensor may have the 20mm range that the stick doesn't measure, but that stick has a muuuuuuch bigger sensor. Obviously aimed at a higher resolution, but I feel the stick is a much better candidate for placing outside the tub looking at moisture, than the little round switch.

 

[Ca++]

Fail on the round one for moisture. Positioned at the tipping point of switching, It could see my hand get within 5cm of it. Quite a good range. Moving away, it didn't think my hand gone until 15cm away. Quite some range. The problem is, the sensors idea of when to say yes, then when to say no again. The distance between these two decisions is fixed. In reality, to see wet is no issue, but just when to let go of that idea, is some way from wet. This thing can see my table, which isn't wet at all, when the sensitivity is high enough to see my hand so far away. It really must be turned down, not because of range issues, but the switching differential. Which is fixed.
This suggests the sticks can go outside the container, as does this presence sensor. As with the sticks, you decide on the differential. However, a tightly defined window between wet&dry, at any increasing distance, brings increasing likelihood of outside interference.

 

[Ca++]

If you understood the last post, skip this one. I don't think I was very clear though (it was late)

Your heater, set to 25c, doesn't come on at 24.99 and go off at 25.00 as it would be on&off at the slightest breeze. Just turning on, would turn itself off again. Long before the space warmed.
Instead, it lets the room drop to say 24.7 and stays on till 25.3 Giving a dead band of 0.6c where the temp can float about, without the heater responding.

This presence sensor can be adjusted high enough to see the wet soil. It also has a dead band to stop it switching like crazy. As the soil dries, it can't dry enough to get out of the dead band. Even crispy soil can be seen as present. The sensor does internally register it got drier, but won't change state as the difference between wet soil and dry soil isn't great enough to move out the dead band. In it's native app, water comes and goes completely, but the soil stays. Which is can still see. However, turn the power off and back on, while in the dead band, and the sensor is off again until it see's full wet again.

The probe type I first sketched out uses a board that sets both the control voltage to switch at, and a control to adjust the dead band. The later board I used, is actually for battery charge protection. So may also work as needed. I think either is infinately easier to work with than programming a mircocontroller with a PC during the grow. When I get a few quid spare, I will have a look. Though the automation thing has my interest now, and is on Amazon for £75 if people like buying from China and paying someone else, so as to not think they are buying from China. Which is much more common, than people actually functioning properly

 

[Ca++]

The fancy pants controller still has my eye.
Many brands sell it, from $50 to $300 depending on store. Amazon UK have one for £75

The PH-W2623 is the EC version, with 0.000mS resolution.
The PH-W2625 is salinity
The PH-W2626 is TDS

The last two will both read in PPM, making them disagreeable. They literally won't agree. Luckily the EC meter is generally the cheapest, and gives better resolution, using standard index units. There is no reason to outcast yourself with a PPM meter, but they are there, if you really want to be a lone wolf. Just don't try sharing PPM readings with other people, as there is a good chance your meter is different to theirs. EC is the common denominator.

One thing bothers me. The probes used in the pics, are all over the place. It seems the designers packaged them in ABS shells with good tank wall suckers. Then the production units get the minimum possible. With old skool glass bodied pH electrodes, an EC meter heads, that have no body. Just wires to dangle them. The EC part is fixed wiring 'for life' While the pH electrode is about a tenner.

£50+damages is pretty cheap for a due meter with online presence. Even if the dosing part was a flop. Though it's just £35+ if you don't want the dosing part. That's cheaper than a new tip for my pH meter.
The video says it all. It's proper bargain kit, sending your phone texts if things go wrong and such.

 

[kokolosh]

I will try this. Anyone use it? Or know about it?

Arduino Automatic Smart Plant Watering Kit 2.1

Based on Arduino Leonardo, the Smart Watering Kit monitors the conditions of your plants with automatic watering system.

www.elecrow.com

 

[Ca++]

I will try this. Anyone use it? Or know about it?

Arduino Automatic Smart Plant Watering Kit 2.1

Based on Arduino Leonardo, the Smart Watering Kit monitors the conditions of your plants with automatic watering system.

www.elecrow.com

The software is on their wiki, they say. It would be worth downloading it, if possible, to get a look at how it works. The fact the software isn't seen anywhere in the advert, means it might not be very good.

 

[goingrey]

The software is on their wiki, they say. It would be worth downloading it, if possible, to get a look at how it works. The fact the software isn't seen anywhere in the advert, means it might not be very good.

There is no "software" in the way you are thinking of it.

It is an Arduino board. It runs by itself. On the PC you just program the code and send it over.

Here is a snippet from their example code:

  if (moisture3_value < 30)
  {
    digitalWrite(relay3, HIGH);
    relay3_state_flag = 1;
    delay(50);
    if (pump_state_flag == 0)
    {
      digitalWrite(pump, HIGH);
      pump_state_flag = 1;
      delay(50);
    }
  }

In this example, if the moisture reading on sensor 3 goes below 30, it sets the pump on. Later there is another test where once it goes over 55 the pump is turned off.

If you want it to dry more between waterings, you change the 30 to a lower number in the code, and then send the new code over..

The software the code is edited with and sent from (Arduino IDE 2.0) looks like this:

If this is "too much", then these Arduino based solutions are not for you.