My Automatic Water Changer - Setup

**SuRje1976** · 09-08-2006, 02:58 PM

Alright.

This is all going to sound a bit insane, and I know I'm going to take some heat for it, because it really isn't that much effort involved in doing a weekly water change. The way I justify it is, now I have an extra hour a week to spend with my wife and daughter. I needed to prove to myself that it could work. What I realized is that there are quite a few ways to skin a cat.

I started toying with the idea a few months ago, absorbed a great deal of info from many of the brilliant people on this forum and came up with a plan. I set it up, and have pretty much changed everything since. I’ve waited to post until I was satisfied with the “final” setup and made sure everything worked as intended. Use the links to see the parts if you’d like. They are not actual photos, but links to the manufacturer's website or to where I purchased them. I intend to update this in the near future with some actual photos. Here goes:

I'm using a Pinnacle+ RO/DI as the water supply. The filter is plumbed directly into my water supply so there is no need to fuss with a faucet adapter. Those things aggravate my wife.

I've got the output T'd so that one line goes to the tank, and the other is available to use for water around the house. Seems I used this line mostly to fill the coffee maker (which happened pretty frequently during the planning and setup). The line runs through several cabinets, behind a refrigerator, through a wall and to the tank.

The waste water exits the tank through another T that was plumbed into the canister's output side. It runs back through the wall, behind the refrigerator, through the cabinets and down a waste line that was also "hard" plumbed. The waste line also accepts waste water from the RO/DI unit. There are check valves everywhere needed.

So here's how it all works:

Approximately 4 gallons of water are drained from the tank daily. Overnight, an electric solenoid (Hayward) and water pump (Eheim 1250) that are plumbed into the waste line, kick on at the same time via a timer. When the solenoid receives power, it opens, allowing water to run through the waste line and down the drain. The water pump is there to speed the process along. It also helps to minimize the variance that would occur with waste water flow rate due to the flow rate variance from a freshly maintained canister vs. one that was in need of maintenance. Draining the desired amount of water takes 19 minutes. The timer then cuts power to the pump and the solenoid (which closes without power). That completes the draining procedure.

To control the water flow back into the tank, I'm using a SpectraPure LLC-S. This is a very nifty device. It is intended to be used to control liquid level in a sump, or as part of a top off system. It's basically a valve that controls the water supply to the RO Membrane (AND the power to a booster pump if you need it to), by the level of water in the tank. Here's how it works:

A sensor (which is basically a hollow tube) is attached to an airline. In my case, the airline follows the waste line through the wall, behind the refrigerator, through the cabinets to the valve which is plumbed inline on the RO/DI unit between the carbon filter and the RO Membrane. I'm not too sure why it needs to go here, but the instructions were very specific and stated that this is where it needs to be.

The sensor, when submerged in water builds pressure up in the airline the deeper into the water it gets. The manufacturer recommends that no longer than 10 feet of airline tubing should separate the sensor from the valve because the longer the span, the larger the distance the water needs to fall before the valve opens. I'm running about 25 feet of airline, and this actually works out to my advantage. It ensures that the water is almost done draining completely before the new water starts flowing into the tank. Had I used the recommended length only, the feed water would be running longer at the same time as the waste line was open and pumping water out of the tank. When the water reaches a certain level (which you must determine, because it varies with the length of airline tubing used) pressure builds up high enough to switch the valve off, cutting the power to the booster pump, and the water to the RO membrane, and thus into the tank. It will not turn on again until the water level drops sufficiently to remove pressure on the sensor (in my case, about 2 inches). When pressure is removed, the valve opens, the booster pump is switched on and water flows into the tank. It turns off when it reaches the set point of the sensor (the top of the tank). I have figured out a way to alter the distance between the start and stop point on the filling sequence if necessary. All one would need to do (I lucked out) would be to fabricate a new sensor (just a tube attached to the airline, remember) of a different diameter. The larger the diameter of the sensor, the quicker pressure builds on the valve switch and the smaller the distance between the start and stop point would be (i.e., narrow sensor = larger water change, wide sensor = smaller water change). I hope this makes sense. I’ve shot for about 10% daily, which approximates the 50% weekly water change my tank was previously accustomed to. I do NOT need to switch my canister off during the procedure because the water level doesn't drop below the intake.

A note - I opted NOT to use the optional safety float valve, BUT I did put a dab of aquarium silicone under each suction cup holding the sensor in place. If that thing were to get free and float, I could wake up to a very wet living room floor! An additional safety feature I've installed was yet another timer that supplies power to the valve controlling the supply of water to the tank. This "safety" timer cuts power to the valve a few minutes after the filling sequence should be finished. This valve, which is also closed without power, CANNOT continue to allow water to flow EVEN if the sensor does not trigger the shutoff, because it won't continue to receive power for very long. Pretty neat.

The entire procedure, from draining through refilling takes just under 1 hour. The drain timer initiates the procedure at 1:30am and supplies power for 20 minutes. As an added bonus I get some overnight aeration and surface film elimination because the water level drops below the lily outflow. Power is cut to the "safety" timer at 2:30am to prevent some freakish overflow. I would surmise that this “safety” could also be used to prevent the feed water from starting until the drain sequence is complete by simply not supplying power to the RO valve until the draining has completed. Like I said, I got very lucky, and the feed doesn’t kick on until about a minute before draining is complete.

I’ve thought about starting the drain at 1:30am, letting it run for about 15 minutes, which would NOT be sufficient to initiate the filling sequence. I would then continue the drain for an additional 5 minutes at 7am. This WOULD initiate filling, and would also have allowed for a much longer period of aeration and surface film removal. It also makes a pleasant trickling sound that, unless I fall asleep on the couch is unlikely to induce nocturnal enuresis. I’ll try it soon.

Now for dosing, I'm using a Rondomatic 400 Autofeeder. It consists of 28 compartments that rotate and tip their contents at predetermined times. It can dose many times daily. Mine doses once, right after the drain sequence. It drops its (dry) contents over the outflow, and the powder gets blown all around. This helps with the solubility of the Equilibrium. GH is dosed one day and KNO3 and KH2PO4 are dosed the other at slightly modified EI (to account for every other day dosing, instead of three times weekly). Iron and traces are dosed 50/50 daily via an Eheim Liquidoser set to max output (~8ml/day).

I also needed to make a small modification to the feeder. I found through reading that I might run into an issue with the contents of the compartments not dropping completely. I did have this problem initially, so I re-mounted the feeder on an angle. The compartment that's dropping is tilted toward the water and everything falls out.

Edit: The Rondomatic did not work out in the long-run. See post #59 for more info!

So, at this point my maintenance is limited to cleaning the glass and filling a Liquidoser once weekly, trimming every so often, filling the auto-feeders (one for fish, one for plants) once every four weeks, filter cleaning and CO2 refilling as needed.

Did I miss anything?

Mori · 09-08-2006, 03:29 PM

I need a few minutes to absorb this, but I like it. Are you using RO because it's convenient to plumb? Or so you don't have to use dechlor?

fresh_lynny · 09-08-2006, 04:58 PM

nice set up and explanation!

steve5520 · 09-08-2006, 06:58 PM

"This is all going to sound a bit insane, and I know I'm going to take some heat for it, because it really isn't that much effort involved in doing a weekly water change. The way I justify it is, now I have an extra hour a week to spend with my wife and daughter. I needed to prove to myself that it could work. What I realized is that there are quite a few ways to skin a cat."

The way I looked at it, being automated, I no longer have to REMEMBER to do it. Or make sure I have the time.

My setup is nearly identical, with 2 exceptions. First, I have a $40 solenoid on the drain line. I go down to 3/8 tube for the drain, so it takes much longer to drain enough water out of my 150 gal tank. I timed filling a gallon jug and did some quick multiplication.
Second, I don't use RO, so the solenoid for that is just hooked to the water supply (just like an icemaker). That's controlled by a float valve in the tank. So I kill two birds at the same time. Autotopoff and water change. The drain solenoid is on a timer, right now I do once a week. I'm considering a little less per change, but going twice a week.

Dechlor? Don't know what to say. I've had freshwater tanks for many years and never used it. Haven't seen problems yet.
And the way mine is now, it takes about 2 hours to change 40 gallons, so the new water is going in farily slowly. I believe chlorine will disapate on it's own fairly quickly.

Betowess · 09-08-2006, 07:59 PM

Oh, fun to read and re-read and re-re-read and contemplate!

**SuRje1976** · 09-09-2006, 01:44 AM

Quote:

Originally Posted by Mori

Are you using RO because it's convenient to plumb? Or so you don't have to use dechlor?

No, actually it would be easier for me to dose dechlor via liquidoser than to reconstitute to desired GH. The plumbing could be done identically for regular tap water as well. I'm using RO because my plants seem to like a very low KH, and the municipal water available to me has a KH of about 3-4. Check this out if you have time:

http://www.plantedtank.net/forums/ge...h-without.html

**SuRje1976** · 09-09-2006, 01:49 AM

Quote:

Originally Posted by Brilliant

so I am unplugging the solenoid after I use the water then plugging it in day later to make more

You plug your current timer into a timer that was set for 12 hours on, 12 off - that would give you every other day as long as there is not battery backup in your current timer. It would also be quite large I suppose. That said, X10 is about as flexible as it gets!

Betowess · 09-09-2006, 02:34 AM

Sergio, why do you need an electronic valve into the drain line, after the pump? Is it to keep any drain water from approaching the pump when the pump/output water is not being operated? Must be. This is definately going to be one of my next BIG projects! Luckily my fishroom is right next to our utility room and I think I can get by without the RO setup.

Also, that is a fairly spendy solenoid. What is the advantage? Is it more reliable than others...why did you opt for that particular unit?

Last query. What kind of timers are you using... Digital?
TIA and thanks for posting. Very informative.

**SuRje1976** · 09-09-2006, 02:54 AM

Quote:

Originally Posted by Betowess

Sergio, why do you need an electronic valve into the drain line, after the pump....

Also, that is a fairly spendy solenoid. What is the advantage? Is it more reliable than others...why did you opt for that particular unit? TIA and thanks for posting. Very informative.

Bob,

The solenoid is there because the pump is not capable of stopping the forward flow of water to the drain line when it is off. If it wasn't there, the tank would be constantly draining because it is plumbed to the canister. I initially was going to drain directly from the tank via the pump, but then I'd have an issue with the siphon that would be started. I did come up with a solution for that, but opted to plumb to the canister so there was one less item in the tank.

I think the solenoid was overkill. My original justification was the 3/8" opening it provided. This prevents excessive back-pressure on the pump. Steve (scolley) pointed out that I could have run a few MUCH less expensive RO valves in parallel.

Don't tell my wife!

**scolley** · 09-09-2006, 03:01 AM

Sergio - Congrats on a very nice setup! Something that can make everyone's life easier to be sure.

But it has certainly peaked my curiosity...

As you know, I've been working on something dramatically similar. So my apologies in advance for asking some rather direct questions. But here goes...

How accurate have you found that sensor to be?

While my objective in creating something similar was a bit different from yours (detecting pressure from the bottom of the tank vs. sensors hung over the side), I spent a good bit of time researching air pressure senors like yours. I even bought two different ones for testing, though not the Spectrapure, which is at the top of the price range for such items.

Have you found the readings to remain consistent over time?

What's the variance day to day? How many weeks have you tested? I found the variance over time (multiple weeks) for similar air pressure sensors to be quite significant. I always suspected that's why Spectrapure sold the accompanying "safety" float valve.

If it works, why does Spectraput supply a safety?

I ripped my air pressure sensors apart to get to the hardware, then grilled the manufacturers about their specs. My findings weren't good. Hopefully Spectrapure is investing in better senors than the products I looked at.

Please let us know if this solution is accurate over time, or has to be regularly re-calibrated. Thanks.

steve5520 · 09-09-2006, 03:10 AM

"I could have run a few MUCH less expensive RO valves in parallel."

I had never thought of running parallel valves. It would certainly drain faster than it is now.

On the other hand, the way mine is setup, I guess I'm really not concerned with draining rapidly. The way it is now, it drains for a bit, then the float valve kicks in and it refills. The whole time, the drain is going, the topoff (fresh water) is kicking on and off.

I have a digital timer set for just under 2 hours, once a week. It takes somewhere around 2 1/2 minutes to drain a gallon. I'm sure the amount draining will vary some depending on flow through the filter, but I'm not sure how drastic it will be.

I guess I figure too, make the error to the high side, change more than I think I should. That way it will probably be enough.

**SuRje1976** · 09-09-2006, 03:29 AM

Steve, thanks for the kind words!

Quote: