[Valentin] is an engineering student and hobbyist gardener. He was planning on going home for a 3 week semester break and certainly could not leave his balcony plants to fend for themselves. The clearly obvious solution was to make an automated watering system!
The most interesting part of this build is the valve. Anyone could have bought an off-the-shelf solenoid valve, not [Valentin], he designed his own. It is simple and just pinches the water supply tube to stop the flow of water coming from the elevated 20-liter water container. The ‘pinching’ arm is raised and lowered by an RC Car servo. When the valve is in the closed position, the servo does not need to continually apply pressure, the servo is powered down and the valve stays closed. This works because when the valve is closed, all forces are acting in a strictly radial direction on the servo’s drive disk. Since there is no rotation force, the drive disk does not rotate and the valve stays closed.
The servo is controlled by a microcontroller. Instead of rotating the servo to a certain degree, the servo rotates until it hits a limit switch. Those limit switches tell the microcontroller that the valve is either in the open or closed position. You must be asking yourself ‘what happens if the limit switch fails and the servo wants to keep rotating?’ [Valentin] thought of that too and has his code measure how long it is taking to reach the limit switch. If that time takes too long, the servo is powered down.
Video below.
I like it, simple and effective.
One other potential way of doing it would be based on weight.
Pot too light? Open valve, add water.
Pot too heavy? Close valve, no flow.
But then you need a precise spring/mass system…
You could also measure temperature, or soil conductivity. But for houseplants just watering every so often should be fine, it’s how most people do it manually.
Anyway, yeah, clever valve! Reminds me of a peristaltic pump. With a peristaltic pump you don’t need a valve to stop the flow, so perhaps that can be version 2, which wouldn’t require gravity to provide the water pressure.
Peristaltic pumps are essentially just a series of radially oriented pinch valves / check valves.
Well, with a little push too. There’s one draining water from the aircon in a place I know, it’s just a motorised wheel with a bump on it (or you might call it a “cam”) pressing on a squashy tube arranged in a circle round it.
and account for the fact that your plant is growing and gaining mass.
Anyone could have bought an off-the-shelf pinch valve instead, too?
Or a pump and use gravity in the other direction.
Plenty of people do, it’s something I often see mentioned on just-buy-the-damn-thingaday.com
This is probably my favorite Hackaday comment ever. Thank you.
Ah the wonderful world of “making”, arduinos and raspis. Build a precarious contraption that has multiple points of failure instead of buying an electrovalve that will last for years or even decades. This is one of the reasons why I visit HaD less and less often.
He was going on break for a few weeks and didn’t need something that will last years…
One day you might find yourself needing to shut off the flow of something else. Like humid CO2, with possible other contaminants. Which destroys the seals in every solenoid valve you’ve tried in months, if not mere weeks. And after throwing away several expensive valves, you might just remember a silly little contraption from HAD. Where nothing is exposed to the damaging gas except some cheap and easily replaceable tubing. It’s cheaper than a commercial pinch valve. And built of parts which can be recycled into other projects, should you no longer need the valve.
The only problem I see here is your lack of imagination.
It’s great that this can be built and knowing how to is an important skill, just don’t forget the opportunity cost involved in building it vs making it and the relative build quality differences.
“the wonderful world of “making”” Then that’s the intended scope of Hackaday and numerous other places on the web, correct? Hell the is the scope of at least a 100 years of print publications. Yea the chances of failure are 50/50, the same odds that we have every second that will live to the next second; thanks to evolution, or some supernatural Creator, if that’s the direction you take.
if you have only a 50% chance of living to the next second, it is wildly improbable that any of us would make it to being 1 minute old.
You live / you’re dead. That’s 50:50!
This is a nice idea but still overly complicated. Why not just make the servo twist the tube to block the water? You would not need any hinges that way.
More wear and physical stress on the tube.
I wanted to make an auto watering system for a long time but wouldn’t think of a good way to turn the water on and off. I wanted it hooked up to the main supply and not a gravity fed container.
Anyway recently I came across these 12v solenoid water valves – https://www.sparkfun.com/products/10456
You can get them in plastic or metal and of course many cheaper sources than Sparkfun.
Just a heads up for anyone else thinking of making such a system!
I think the comments are centered around whether a home-brewed system like this would be durable – the worst that could happen is that it doesn’t work (either dumping all the water at once, or not operating at all) and for balcony plants that’s just fine. Personally I’d have had the blade pull down between two stops to put a double-kink in the tubing, but that’s just me.
As far as buying an off-the-shelf 12V valve to use with mains water, the local home improvement stores will have durable, outdoor-use valves in the sprinkler section for ~$12 – $20 and I’d trust them a lot more than something intended for appliance use for half that on Ebay, particularly if any of it were going to be indoors.
Solenoid valves that they sell for sprinkler systems tend to be the kind that need a decent amount of positive pressure to operate. The solenoid part only opens up a bypass that allows the pressure behind a diaphram to drop and the water main pressure pushes it open. Good for high flow, bad for low pressure and this application.
I think controlling water from the main supply with a valve like is a good way to do it for outdoor watering, but not so good indoors. Insurance will pay for damage if a valve in a dishwasher or refrigerator ice-maker malfunctions, but doubtfully for a home-built watering system! I built a house-plant watering system last year, and what I did was use a pump to pump water from a 5-gallon bucket on the floor into a small elevated reservoir with a liquid level sensor, and then released water from there to different plants with gravity-feed solenoid valves. This gave both a way of metering the water, and a level of safety: the total possible water release from a software bug was the contents of the bucket, and from a valve malfunction or tubing leak only a few cups/deciliters.
I don’t think I would would have thought of using that method of controlling the water. Most likely I would have opted for a larger or multiple reservoirs, and an extremely low drip rate. Having linkage that breaks over center to lock something is very common. Locking pliers and other clamps. Drive line parking brake on l medium duty trucks or other trucks not equipped with air brakes.
I would recommend shaping a radius on the edge of that pinch bar.
The corners it appears to still have will be pretty good stress raisers
and likely be the source of a more abrupt failure of the tubing
in a perhaps unexpectedly time frame.
Second thing is I would put just the slightest offset or small notch
in the upper surface of the slot.
Or a curve ( think of a frown shape, as the bar is currently mounted)
So that the chances of the wheel drift or a slight shift in the motors stopping point are reduced of allowing the valve to creep open and leak.
You may now resume the” HackaDay is headed to Hades in ‘duino box” discussions. ;^)
This is a hack. It’s an ugly, effective, and deliberately short term. It doesn’t have to be efficient or elegant it just has to work.
In answer to the question posed by the creator in his video. I would have used a water globe. Maybe something like these cheap DIY ones if I didn’t already have the fancy glass ones on hand. http://www.youtube.com/watch?v=5d2LjY2P47s
Thanks Leithoa! Thanks for the link, too! And yes, it was an ugly outcome of an afternoon of hacking and tinkering trying to solve a problem within one day and only on-hands-parts – at least it is moving and a little bit satisfying to watch! :)
NIce job. A solution that will do in a pinch.
[groan]
Would be nice to consider the readers a little . Don’t start videos automatically . Some people don’t have a huge download capability.
I was interested in the code but could not find a download link anywhere
This is gravity fed, right? Why not lower and raise the the other end of the siphon in or out of the water tank with a motor, pulley and a string? Almost no wear on anything.
You need a motor that can lift 44lbs(20kg) vs a servo that needs a few inch/lbs of torque to crush some tubing. If you only have one or two plants you could use a smaller reservoir in turn allowing you to use a smaller motor and making your solution viable, but the creators needed a large volume reservoir.
*inch-lb
You’re right about needing to lift 44lbs, if you need to lift the entire reservoir. Since this is for a smaller set of plants, you could have the water reservoir on the ground with a smaller vessel attached to a motor, and have that smaller vessel drain into the plants.
When the motor lowers the vessel into the reservoir, the vessel refills through an opening.
Still going to take more power than a servo, but if you’re working with more than one or two plants, that might be the way to go.
If you raise the tube out of the water then all the water will drain out of the tube. How would you restart the siphon for the next watering?
If he’ed put a knife on it he could slice cheese with it too.
I used the following items to build an automated patio-garden irrigation system.
20 Gallon storage container ~$6
Swiftech computer water cooling pump (MCP655) ~$70 7 years ago.
An old 2 amp 12 volt DC power inverter. Cost was free since I don’t even have the device it went with anymore.
A cheap 7 day / 8 program digital outlet timer ~$9
A rainbird drip irrigation starter kit with 25ft 1/4″ line, 50ft 1/2″ line, and various elbows, connectors, and spouts. ~$23
A pack of 50 adjustable drip spouts. ~$5
The whole system is controlled by the digital timer. Set to run for 6 minutes, 8 times a day (every 3 hours). The timer activates the outlet, which sends power to the DC inverter, turning the pump on. The pump pulls water out of the tote and pushes it through the drip system. The components cost me < $40 out of pocket due to some items laying around such as the CPU water pump, storage container, and DC inverter. But even if purchasing those items, the whole system would cost < $100 with some searching for thrifty prices.
I've got it set up to water about 8 potted plants, but it could be expanded indefinitely to however many plants I could fit on my patio. I live in Florida, so the summers can get hot, which is why I have my system set up to water so frequently for short periods of time. The rate of flow of water with my current configuration would drain my 20 gallon tote in about 5 days, so I've been considering upgrading to a larger tote if I decide to take a longer vacation than that.
A possible future upgrade would be to add a wifi outlet timer, like the Belkin WeMo, so that I could control it remotely. For instance, I could shut it off on days that it rains heavily. Or set it up with IFTTT support and adjust the timing based on any weather conditions.
You could probably just wire this up to a timer and call it a day!
http://www.electricsolenoidvalves.com/1-4-120v-ac-electric-plastic-solenoid-valve/
The whole project would probably cost you less than $40 if you don’t have the spare parts laying around already.