Had to happen. Because I don't have enough stuff I want to do...
I want to automate the garden. I've worked to move the vege garden and I'd really like to idiot-proof (read "me-proof") it because some days the spoons are in short supply and I don't have the energy to water - and in this climate that's a disaster, yellow leaves, stunted growth, if it happens twice in a row we lose a crop of something.
OTOH, things here get waterlogged really quickly and that can drown plants too. We end up with a lake at one end of the garden. And yep, even the reticulation can do that. Ask me how I know that. 😸 (Spoiler: The temporary mechanical timer I was using at the time got stuck on. For two hours.)
 |
| Yeah. A lake. That's a rubber boot I'm wearing. |
Being a bit geeky has advantages. I decided to program up an ESP8266 (D1Mini) as my retic controller, and because I know stuff, I gave it three circuits to manage, and timed them so they run A-B-C-A-B-A in short bursts so that circuit A runs three bursts during the night's run, B gets two bursts, and C gets just one squirt.
My rationale is that when I water anything by hand, I notice that the first time I water for the evening, water pools but doesn't really spread, so I've always been in the habit of going round twice, and the second time the water soaks in properly. I.e. several short waterings is better than one longer one.
Caveat: Once soil is conditioned and has plenty of organic matter in it, water "wicks" sideways just fine and you may be wondering what I'm going on about. But before the soil develops the organic component - and, more importantly, after a few days where I hadn't for various reasons been able to water - it develops some water repellant properties. I theorise that the microbes near the surface get killed by heat and their wax does this.
Solution: Mulch is the key, but it's not always everywhere. Again - spoons, priorities. Also the only mulcher I've been able to get hold of was a complete waste of the money and time I put into trying to recondition it and has since gone on to become spare parts.
Anyhow - several smaller waterings in succession works better for me. It's probably better for most gardens but a pain to do with off-the-shelf reticulation controllers and so I'm making my own. And it means I save water, instead of watering everything for 30 minutes and having water pool everywhere and dry spots, I'll water three separate circuits for an hour, get the right amount of water for each situation (and consume only 1/3 as much water for double the time, saving 1/3 of what would have just been runoff in one watering) and as I'm currently doing this manually using my phone clock and three inline taps, I've already seen how well stuff thrives under this regime.
Hence, watering in shorter bursts. It works really well. Ten minutes on, then off while the next circuit gets a burst, etc. But then a thought hit me. What if I have this all automated - and it's rained that day? I could just put a simple moisture type rain sensor out and latch an input if it got wet that day. (But it wouldn't tell me how much rain had fallen. Hmm... )
Also, checking soil moisture wouldn't hurt. And I could log that stuff and also monitor how much water goes through the system. Temperature. Humidity, as that affects transpiration rates.
Insolation. If I was going to power this with a solar panel (and who wouldn't, given they already HAD a largeish solar panel and lead acid battery powering other stuff, and that kind of data directly affects plant growth) I could measure how much sunlight we'd get here day by day by season by season.
Feature creep was in full swing. . .
I decided to ditch the D1Mini and got a NodeMCU Nano form-factor thing and that's arriving in a few weeks. The board is very affordable at Aliexpress. And I've always been a bit of a data wonk, the more I/O pins you have, the more data you can collect, the better you can make things work. And - rainfall figures. . . THAT'S what I'd been thinking about! There are tipping bucket rain sensors out there that can tell me how much rain's fallen!
Tipping bucket rain sensors are cool. A little inverted "T" shaped toggle sits under a funnel with a known surface area. One side of the T fills with water because it always falls to the left or the right, that side fills up - and it tips because that side now has a few ml of water in it. As it tips, it falls to that side, empties, and stays that way because it's overbalanced. And now the middle upright of the T diverts the dripping incoming water to the other side, until it fills up, overbalances to that side, and the process repeats. Along the way, you trip a switch of some kind or ratchet a counter mechanism. In my case, you whiz a magnet past a Hall effect sensor and count pulses, then add them up in the MCU.
Tipping bucket rain sensors are cool. Until you go to the least expensive online site you know of, type in "Tipping bucket rain sensor" - and . . . Expletive deleted . . . When that one sensor can cost more than the rest of the retic system including the solenoids and plumbing, unaffordable, I have to do without it. But. . . I have a 3D printer. And people have made these things, there are models online on all the major model collection sites.
And yet. . . I'm a masochist, I have to DIY. A couple of hours in Tinkercad and I have my own version. The only bits I really need to print are the bits inside the box. The tipping bucket, mounting, and parts to hold a small neodymium magnet and Hall effect sensor. The box around it and the funnel can be anything at all.
Here's the secret to calibrating it:
Here's how rainfall is measured. You make a straight-sided container, and the depth of water that collects in the container (in millimetres) is the total rainfall. For any size container as long as the sides are straight and parallel.
With a gauge, you look at the surface area of the device, work out how many millilitres are required to produce one millimetre depth of water in it, and then test your tipping buckets to see how many clicks it would take to measure that much water slowly poured into it. Which, if I use a 10cm x 10cm square funnel, will need 10ml to form a 1ml depth of water and I am before having actually printed it, estimating 4ml per bucket so I'll be dividing my clicks by 2.5 to get a reading of 1millimetre depth.
The point is to know to within a tenth of a millilitre how much water goes through your particular buckets, because then you can put it under any sized collector and work out what you have to divide your total clicks by to get to the millimetres figure once you know the area. You can even put the funnel out somewhere and run some tubing back to your counter. If you do that, remember to provide a way out for the water once it's been through the tipping buckets!
Here are some of the elements of the reticulation system:
Firstly, I'm only using drip hoses and drip nozzles that produce a tiny low spray around them. That way I can pile some mulch over the drip hoses to keep the water out of the wind and sun, and the drip nozzles can (hopefully) stick up that bit more and spray over the mulch, spreading the water farther. All those sprays and mist sprays are prone to two things, wind drift and evaporation. Fine sprays like that are perfect inside greenhouses and long grow polytunnels but they waste half your water budget due to those two failings.
 |
| Img1: Description below. |
Quick description:
- A - These are the tips I'm standardising on for almost every purpose, and also this 4mm tubing. (As you can see, prices are good compared to our good ole Aussie hardware stores.)
- B - shows the circle one of these can water. (Hand for scale.)
- C - this is about the lowest setting I can get from them, handy when watering a small pot, but also slightly hard to get right so I may replace the ones in small pots with a more regulatable version.
- D - this is a similar tip that's easier to regulate (but not by much) and that I bought a few years ago and am trialling. I think they'll be replaced though.
- E - is one of the weep holes in the weeping hose. They're spaced 300mm apart and produce a steady drip over a wide range of pressure.
- F - one other thing, if you can afford it, putting these taps everywhere can be a bonus. They allow you to turn a small section (one raised bed or one series of pots) off to conserve water while you're working on them or even do some really rough water regulation by partly turning them off.
The background is three raised beds and several pots around the south end of the garden after winter but before spring cleanup. (You're looking for the 3rd raised bed, arent you? It's at the extreme right behind the rush screen...)
- G - One thing not shown is the watering into the worm feeder tubes.
You should probably refer to this older post about point G above. I made some worm feeders that go into each raised garden bed, and they're just 100mm PVC pipe stuck about 30-40cm into the ground with a lid on them. There are holes in the side down the bottom tolet worms in and out, and I toss compost, clippings, prunings, and other worm food down there, then put the lid back on. Worms wander in and out and eat their fill, then wander out - hopefully to the vegetable roots - and leave their nutritious castings (aka worm poo) there for the plants. In the process they also aerate the soil and allow water to penetrate better.
That was fine but the tubes needed to be opened too many times to add water to keep the worms' tucker moist, and they hate being disturbed. So I used the same line as feeds all the vegetables in pots (they get one or two bursts of watering each night) and drilled a tiny hole in each lid, poked some 4mm tubing inside, and put a red dripper in. Now I can just top up the food (once every few weeks/months) and the rest of the time they have peace and safety down there.
Footer
As always, please share this article and my others like it, go to my News Stand to see all my other posts and share links to the News Stand and any articles you found interesting, and if you can, donate here or here and find out here why it's important. Or subscribe to my once a week newsletter and stay in the loop.
No comments:
Post a Comment