Redundant Automated Water Filler For Your Coffee

We’ve always wondered why we have indoor plumbing if it isn’t hooked up to our coffee pots. We probably drink as much coffee as water anyway, so why not just hook up a water line to refill the pot? [Loose Cannon] aka [LC] has been working on just that problem, with a whole lot of extra features, creating a very robust automatically-filled, gravity-fed, vacuum-sealed water tank for whatever appliance you have that could use it, including your coffee pot.

[LC] tapped into the 1/4″ water line from the ice maker, which has the added bonus of being a common size for solenoid valves. He’s using an eTape sensor to measure the water level in the reservoir, but he ALSO is using a flow meter in the line itself to double-check that the reservoir won’t overflow. The flow meter allows a hard limit to be set for the maximum amount of water allowed into the tank. He’s used an Arduino Micro to tie the project together, which also handles a real-time clock so the tank can be filled on a schedule.

The tank that [LC] was trying to fill was vacuum-sealed as well, which made things a little trickier. Without a vacuum on the tank, the water would just run out of the overflow valve. This is an interesting project that goes way beyond the usual automatic water supplies for coffee pots we’ve seen before.

14 thoughts on “Redundant Automated Water Filler For Your Coffee

  1. Honestly? I think i’ve modified every coffee pot I’ve come across for this at every company I’ve worked for. Most people are lazy and refuse to refill it. It’s not that hard to do. There are websites on-line that sell ultrasound transducers that monitor fill levels, and all of the other “fixin’s” that are required.

    More people need to be thinking like this. Our homes would be a lot better. Not to mention easier to use, more efficient. Now to keep the wife from bitching that something was modified, and/or the house was ripped apart for a week.

  2. I’ve thought about this a lot over the years, usually for plant watering systems.

    The thing that always gets me is the reliability. If something goes wrong, the system could dribble water continuously, causing enormous damage to things in the area, possibly structural damage.

    Note that we don’t have automatically-filling tubs either. When the tub overflows, it wrecks the downstairs neighbor’s apartment.

    Also note that standard plumbing doesn’t have error detection either. You can have a leak in a pipe somewhere in your home and you’ll never know before it does major damage. You can have a leaky toilet run up hundreds of dollars in water bills and *never* know it.

    Plumbing engineering descends from the time of the Romans and reflects this fact. A more modern system might have a smart valve at the meter which communicates with all the possible outlets (faucets, toilets, and so on). You turn the faucet on and it communicates with the valve to say “it’s OK if water continuously flows”. You start the tub filling and it says “I’ll need *this* much water to fill”.

    If water is flowing outside of the request/supply model, you know something’s amiss and the smart valve can stop the flow until the problem is addressed.

    This wouldn’t prevent water damage, but it would greatly increase reliability. Instead of relying on the pipes not bursting, you’re relying on the pipes not bursting at the same time that the smart valve system fails.

    I’d love an automatic plant watering system. The thing that stops me is the reliability – I don’t want the dishwasher replacement solenoid valve [from eBay] to stick in the “on” position, causing major structural damage.

    1. That’s the same reason I don’t have an automatic fill on my plant watering tanks. Today I have a short hose at the sink, where I fill gallon jugs to carry to the tanks. The biggest leak I could have today is that a 10 gallon tank could rupture, and I can fix that size of a problem with a mop.

      I also won’t run a long hose to the tanks just to avoid carrying the water. It would take about 15-20 minutes for my reverse-osmosis system to refill them. If I had a long hose and a valve where I could fill them, it would be just too easy to get distracted during the operation, and I’m afraid I’d come back to find the valve was accidentally left open for an hour or a day. Today, when the pressure gets low, it takes 2-3 minutes to fill a gallon jug, and I already find I can get distracted during that time; at least the jug is in the utility sink where the overflowing water safely runs down the drain.

      If I automated the valve and put in an electronic auto-shut-off system, I’m afraid it could still easily fail (what if the breaker tripped while the valve was wide open? What if the home controller crashed or rebooted during that time?)

      Mechanical float-controlled shutoff valves have existed for a long time, and I have friends who have such a system in their water tank. I’ve considered that as a possibility. I already have my home automation system monitoring for floods. But trusting an electronic controller as the last stop before a flooded basement? Not yet.

    2. I do have an automated solution for short weekend trips out of town. I have an old submersible utility pump I put in a 5 gallon bucket, and I have a small drip irrigation system from Home Depot hooked up to it (including the 25 PSI pressure limiter.) The tubing and emitters are placed on each of the plants that needs daily watering. The pump is controlled by the home controller via a Z-wave appliance module. I set it to go on for 30 seconds each morning. 5 gallons is plenty for a weekend trip; if we’re gone for longer, we have so many plants that require a twice weekly watering that we hire a house sitter to take care of both the plants and the dogs.

      If things go horribly wrong, the worst that could happen is 5 gallons of water could spray around the room (and the pump would then run dry for a few days, which is more concerning).

    3. “it would greatly increase reliability”

      Until the power goes down and the batteries die and your “automatic” system becomes non-functional!

      For a tiny fraction of the hardware costs, you can simply purchase homeowner’s insurance that will reimburse you for water damage from burst pipes.

      If your pipes are bursting, it’s because your furnace is down, the temperature has dropped down to freezing and your plants are already dead. You’re better off with a temperature alarm.

    4. Hi there,

      Thanks for your comments, I agree automated water solutions can be treacherous, since there are known unknowns, and we can control for those, but we also have unknown unknowns. My project attempts to control those unknown unknowns in three ways:

      First the solenoid valves don’t open until voltage is applied, it’s true, the solenoids could fail and stay “stuck-open”, and that would ruin my day. However they are only open for very short periods of time and rather infrequently, so I am guessing that the low duty cycle will help with longer life span. Also, I do use a cheap stage 1 inline water filter to make sure I trap any particular matter that may be coming in from the water source.

      Second, I use two methods for calculating water flowed. The first is the eTape that provides the initial “fill” command when the reservoir empties. The second is an inline fluid flow meter that measures how much water has flowed. If the eTape malfunctions and water keeps going, it will shutoff when it reaches the flow meter hard limit. The hard limits are calculated by some live testing to determine accuracy and error range, and then adjust the limit values to account for these variables. Therefore you arrive at a certain max values and compare those to the values measured for the total amount of fluid the reservoir can hold (say 2 liters or 6 liters in my case).

      Third, the 1307 RTC ensures that the system is only enabled for refill when someone is around or when the device is likely to be used. In other words, it runs on a schedule, turning off at night and running during the day.

      Last but not least, here are some known unknowns:

      1. The solenoid valve stays stuck open: solution is to add an external flood monitoring solution that can alert you via SMS or email. You might want this to run on the arduino, but if you were extremely cautious, you might want a completely external or isolated system (non interactive).

      2. The vacuum pump may fail to start: there’s no way currently to tell if a vacuum is being pulled. The current logic tests to make sure that the eTape water level values are rising during filling. If they don’t it could point to an overflow situation, one where the vacuum pump has failed. A secondary way to ensure that this issue is prevent is to install a vacuum sensor, and confirm that a vacuum is being pulled before activating the filling solenoid.

  3. Smarter water control and monitoring could save money and prevent damage.
    With water becoming more a valued resource that we should manager better maybe it’s time to look more at opportunities for improvement.

    1. i might spill a couple of drops of water here and there as I fill my coffee maker from my filtered pitcher, I’m sure the planet’s imminent destruction could be avoided if I were more careful

  4. What happens when the power goes out and you want water from the smart system? If you can’t open the valve to get water for a drink, you’re kinda screwed. We can apply a lot of complexity and unreliability to a system that while dumb is amazingly robust and foolproof. This could end up as a classic case of fixed until it’s broken.

  5. Apparently you haven’t been keeping up on mid-21st century advances in plumbing. Plumbers have recently invented this novel device called a “tee”. This marvel of engineering diverts water from one source to two pipes! That way water can now flow to both pipes at the same time! So if you have an electric valve filling the coffee pot and the electricity fails, you can still use the manual valve at the sink to get a refreshing drink of water. Isn’t that just amazing?

  6. Has anyone ever thought to put a dehumidifier on a coffee pot so the water can be pulled directly from the air? I don’t drink but a few cups of coffee in my home, and am not sure how much water a dehumidifier can produce, but I think it wouldn’t be too much engineering to make it work.

    1. I once drank the water from my parents dehumidifier when I was a teen – I recall that it tasted pretty musty. I realized that dehumidifiers don’t produce sanitary water. Any bacteria, spores, molds, fungi, dust, dust mites, or other air pollutants in the room land on the chiller along with the air, and they can continue living and breeding in the reservoir. Fortunately for me, I don’t recall getting sick from it.

      The dehumidifier water is free of minerals, though, and works as a replacement for distilled water or rainwater if you are watering plants, or filling a steam iron. And I suppose the boiling of the water for brewing the coffee would kill whatever was in the water before, so it could be doable. As for me, I’d rather pour a glass from the reverse osmosis filtered storage tank.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.