Greenhouse irrigation

I'm not a gardener, but my friend solves the watering issue by growing them hydroponically -- no dirt, no digging just a pump to circulate the nutrient solution. I'm helping him with automating the nutrient mix from concentrates. Its on hold until fall, but I have a prototype I am testing.

Most of the automation will be to detect flow stoppages and/or low nutrient levels in the circulating tank, and ultimately refill the circulating tank from concentrate reservoirs and warn when concentrate levels get low.

Its 100 degrees out there today and the plants are still alive, given my "black thumb" seems the way for a lazy man to garden :slight_smile:

1 Like

I used to grow hydroponically, but it was proving much too expensive (in the uk).
The cannabis trade has ramped up the cost of nutrients, ph regulators, clay growing medium, etc.
So just not worth it when growing home produce on a small scale.


1 Like

Do you have only one type of plant (tomato)? If you have more than one variety how do you measure the water to each different type of plant? I would suspect that a cucumber would need more/less water then a tomato...

The dribblers that I'm using have interchangeable heads which deliver 4, 8 or 16 Litres an hour, so although I'm using 16 Lt dribblers on tomatoes & cucumbers, I'm only using 8 Litre dribblers on the chillis - which don't transpire as much.
When the cucumbers were small plants, they too were using 8 Lt's, but as they got bigger, they were put onto 16 Lt's, and they are thriving well, over 2 mtrs tall now!

It's all about balance, and establishing the ratio between different plants needs.

I could always click-in additional dribbler pipes into the feeder pipe (they just push in) and add another dribbler (16 + 4 = 20 Lts), but I haven't found it necessary.

1 Like

I used this model

Yes, a resistive type.
They are very inaccurate, and heavily influenced by the soil temperature and soil nutrients.
I tried those first, and slung them in the bin!

You really need a capacitive sensor. Much more stable, and deliver more consistent readings.

That explains how you know how much water you deliver. That was another puzzle I was curious about. Another puzzle is how you came to the 62% soil moisture number.
How did you determine the "ratio"? It would be nice to have a chart but I suspect it was trial and error.

Trial & error!
I started at 50%, which is mid-way between fully saturated soil & oven dry soil, but having dabbled in gardening for over 50 years, I could feel that the soil was too dry, so gradually increased it until experience told me that the humidity was appropriate. Plants are very forgiving and will give clues if they are not happy.

As above.

I suppose the upshot of this project is that it looks after the plants much better than I usually do, albeit not perfect, but good enough to provide a bumper crop despite the UK heatwave.

1 Like

We are suffering in the heat wave too. I just read that the whole northern hemisphere is in the same boat.
I remember growing up in Northern California Summers that peaked at 120F and now they're saying the record is 118F. Hey! I remember official temperatures higher than that! I really don't want to remember them anymore though. Makes me sweat just thinking about it.
It's to late to start this year but I'm really looking forward to next spring planting and a full winter to work out the bugs in an automatic irrigation system. Thanks for sharing your project.

1 Like

I'm doing this only for the technical aspect to help out my best friend, as chicken is my favorite vegetable :slight_smile:
But getting supplies at the local home&garden store sure looks like it'd always cost more than super-market produce, I'll defer to the arguments that "its better" from people who actually like eating vegetables :smile:

I'm curious as to what is used for "pH regulation" My brother-in-law got really PO'd at his pool guy after I told him the $25/1.5Kg sodium dihydrogen carbonate pH increaser was just "baking soda" -- $7/6Kg at the wholesale club.

Sorry for perhaps being a bit off topic, but I'm getting useful info about dribblers and sensors here. I would think these cheap moisture sensors could be a good alternative to float switches zenofmud mentioned. Seems to me the float sensors I've found need holes in the tank top and bottom, where the resistive sensors could just be mounted in the inside of the tank and the wires routed along with those for the submersible pumps, or do they not stand up to water immersion?

They do but I used a cheep 18 gallon plastic tank and it was easy to put the holes in and wire them up.

@Paul-Reed - is there a spot in the plant's pot you've found is the best place to put the moisture sensor? Thinking aabout it, it seems that the distance from th edribbler will effect the moisture in the soil. On the other hand I could be all wet :grin about that.

I'm using grow bags which house 2 plants each.

The water entry point is midway, between the two plants, and the sensor is inserted through a slit in the side of the bag about 500mm from the water entry point.
This ensures that the water added doesn't flood the sensor, and as it's as far away from the water entry point as possible it ensures that I'm measuring possibly the driest part of the grow bag.

That's also the reason that I allow 10 minutes between readings, to allow water time to soak throughout the growbag and influence the sensor (which it does fairly quickly).

If I was using pots, I would put the sensor at one side of the pot and the dribbler at the other, and also put a generous layer of gravel on top, to ensure that the very top of the soil doesn't dry and crust, making it impervious, and increasing absorption time.



Do you run VCC to your sensors/Wemos or use a battery pack? My setup would be about 18' from the nearest outlet and I think I'll be using battery packs, but how long they would last is an unknown.

hi Paul

thank you for your contribution!
I'm now working on a same project, not for a Greenhouse, but for my garden.
My project is, a Raspi-Pi2 for the central and 3 ESP8266 for the 3 stations
(station 1, Pump with 2 Valves for sprinkler / station2, with 4 Valves / station 3 with 1 Valve)
on the Pi is node-red running and at the moment I have some of troubles to find the right way
my first question, MQTT, I was reading a lot about, but is this necessary only when for external use (outside from the "home-network") or it helps generally?

That is not necessary if using the PID algorithm. In general the faster the algorithm is given data the better the control will be. The algorithm will cope with the transport delay between the input and the sensor, though the tuning setup might be slightly different with the faster sampling. However if there is random noise in the sensor which is being filtered out then there can be a benefit from sampling the sensor more rapidly, filtering it, and feeding it to the algorithm at a slower rate.

I'm working within my home network, and MQTT provides a way that I can maintain some central control over a remote device. You could of course program the ESP to operate independently, but I'm a control freak!

To power the pump as well? I would have thought that 18' of cable would be cheaper and more effective than resorting to batteries??

I see your point but my plan was to place my pump closer to the outlet and run a long length of hose.
I may even have to do some home improvement...
I also have to place my plants on a second story pergola in order to capture enough sun. It would only be accessible by ladder.
This is all in the first stages of planning. :slight_smile:
Have you run a test on how far the wemos can be and still provide reliable data?

Well, my wemos is a generous 15 mts away from my router, and separated by 2 solid walls (the router is in an integral garage, in the house) and I get a good solid reliable signal.
This morning, I'm seeing a RSSI of -65 which is pretty good as it's usually in the low 70's.

However, router performance can greatly influence this. For example, I've recently changed internet provider (TalkTalk to EE).
The router which EE sent me was absolute tosh, and resulted in a RSSI of < -75 and periods of failing connections. That went in the bin, and I'm back using the TalkTalk router again.

Also, the Wemos is directional, and you need to rotate it until you find the strongest signal. I've attached a sketch which I used to locate the optimum Wemos position and direction.
Just upload it, and read the results in your Arduino serial monitor.

rssi.txt (4.1 KB)


Well, in that case if my router is half as good that changes everything. I could reach my wifi with the wemos and keep my RPi inside my office. Sweet.

There are various Apps for Android and iOS that measure WiFi signal strength.
inSSIDer is quite a popular one.
Once installed on your phone/tablet you could walk around the garden and check the signal strength at different ranges.

Cheers from David.