My apologies if this was designed by someone else, I came up with the concept independently. I don’t mean to reinvent the wheel, but I haven’t seen this setup before.
I’m looking to setup a system to do water changes automatically.
~It needs to be reliable, because if the house floods, my wife will kill me.
~I don’t trust float switches / float valves and don’t want that clutter on my tank.
~I don’t trust automatic drain valves: if it gets stuck with a dead fish / weeds / fails open . . .the tank will drain without my knowledge.
~I have very hard water (16-20 degrees general hardness) and want the critters to be comfortable in soft water.
So here’s what I'm putting together:
This is a trickle water-change setup that will blend DI/RO water with tap water. It works without float switches or float valves and is pretty idiot-proof. Because of the solenoid valve placement, there’s no risk of feeding hard water backwards through the RO unit. It should be very reliable because the only vulnerability is the overflow; since the water is trickling in at less than 5 gallons per hour, I’m not too concerned about flooding the house. As for my approach, I argue doing a small water change every day is better for the critters than one big 50% change once a week.
Drilled aquarium with overflow and permanently plumbed to a drain (sewer, garden, etc)
RO/DI unit: 50 to 100 gallons per day output
1/4" polyethylene tubing
Dwyer instruments flow meter: VFB-80-SSV (rotameter with valve to throttle tap water)
Dwyer instruments flow meter: VFB-80-SS (wide open rotameter for RO/DI water)
Dwyer instruments solenoid valve: SBSV-S1F1 (Normally Closed, 120v AC, 188PSI WOG rated)
Miscellaneous 1/4” fittings
Here is the final product:
I just ordered my parts today . . . one of them is a “special order” item ++sigh++ hopefully it’ll show up in a few weeks. Till then I need to hook into my sewer line.
After a bit of discussion on the board, I'm changing where I draw my tap water from. The hardwater tap now is being taken off just in front of the RO membrane. All other aspects of my setup remain unchanged
Here was the original idea:
I have the drain all plumbed in and its working exactly as planed. I put a loop in the drain hose which will serve as a 'trap' and prevent sewer gasses from coming back up the overflow. My only negative discovery is that the manual drain valve is extremely loud!! The gurgling can be heard though out the house. However, the overflow is silent and I can deal with a little noise for the few times I actually need to drain off a large volume of water.
This is the only valve I am installing on the drain side. All parts here are 3/4" (hose barbs, valve, etc)
Here's the aquarium's plumbing
My first box showed up. I got the solenoid valve and rotameter with needle valve.
The little bullet shaped piece of stainless steel is the float that gets pushed up by the flowing water. That indicates how much liquid is passing by.
This is a piece of scrap aluminum we had laying around at work. I cut and bent this into a bracket to hold the rotameters
Here the aluminum is bent into my "control panel"
I got the hardwired timer purchased and installed. The whole build is done, except for the RO rotameter.
I couldn't wait for my parts to show up so I jerry-rigged it around the missing component. I tested for chlorine break-through and even with 2+ year old carbon filters, I'm BDL (below detection limit) on chlorine!!
The timer is acting up: it turned on at 7pm as it should but was still running at 10pm (it was supposed to shut off at 8pm). . . I did a little reprogramming and it should be working correctly now. And according to Dwyer, my other box shipped out yesterday. Monday I should be be able to finish this project.
I got my last component in and everything is now assembled. I "fired it up" this afternoon and everything is working perfectly!!
I didn't discover any design flaws and I don't have any regrets. The only suggestion I have someone wanting to copy this, go for the 75-100 gallon per day RO filter. A higher output filter will run the rotameter more in the center of the range.
Good luck to all of you and be sure to post your DIY's for the rest of us to admire / learn from / laugh at!!
This is the back side of the rotameter "control panel." I just used plastic quick connect fittings. Homedepot had exactly what I needed!?!!
Here is the final assembly with important parts labeled
And this is the system when it is running. When I snapped this pic I was still playing with the flow rates.
Now that the whole project is done, my only suggestion to those looking to copy this is: Go with a 75-100 gallon per day RO/DI filter. My 50 GPD rig barely puts out enough flow to register on the rotameter. Not a biggie. Total price on this thing in 2012 dollars is about $400-500.
$120 for the RO filter
$200 for the rotameters and control valve
$30 for the timer / electrical components
$50 for the tubing, plumbing fittings, and drain parts
$50 for a diamond hole saw and bulk head fitting for the aquarium glass
$50 for shipping, taxes, misc expenses, screw-ups
I'm now almost a month into this automatic water change system. The water chemistry changes are quite dramatic! One area of huge improvement is on phosphates! Up until now, every time I did the test, it was "PEGGED" off the chart at 10+++ PPM. The test would turn ink blue in seconds . . . I didn't need to wait 3 minutes to know I had tons of it. (I think that problem was from overdosing 7.0 Ph buffer)
As of today, I'm testing at:
0 ppm PO4.
0 ppm Nitrate
0.25ppm ammonia (I'm suspicious of this, I don't think that's possible with my moderately-to-heavily planted tank.)
The tank is heavily fertilized with root tabs and all my plants are root feeders. Even so, the plant growth has slowed way down. It seems everything was thriving on my neglect! ++sigh++ we'll see how ambitious I am, but I may end up doing a quasi Estimative Indexing fert schedule.
Today things are looking very healthy. If you look really close, you can see the overflow in the top left corner. [The arrows are for an unrelated thread, but they point out my 3 awesome Nymphoides Aquatica (banana plants)].
Here's a little update...
My setup has been running FLAWLESSLY for over 2 months now! The fish have never been healthier, the water chemistry has never been better and honestly I don't know what to do with myself.
In the past I was always chasing after the water parameters because something was perpetually out-of-whack. I was always fighting with the ph, or the nitrates, or something . . . but now, it just runs on auto pilot.
If you're even remotely interested in doing an automatic water change system, my setup is the way to go!!!! Very simple, very reliable, with fantastic results! As long as I'm into freshwater fish keeping, I will be using this system on every tank.
Today I checked the chemistry and thought I'd post the results:
Feed water from system:
5 deg general hardness
0 ppm nitrate
2ppm phosphate (From adding 7.0 ph buffer)
0.1 ppm(?) amonia
9 deg General hardness
About 1.6 gph of RO water + 1.1 GPH of tap water = 2.7 gph and it runs 1 hour daily.
2.7 gallons * 7 days a week = 18.9 gallons a week.
The tank is 72 gallons, so that means I'm doing 26.3% water change every week. That seems to be just perfect for my stocking of plants & fish.
Sorry for the window reflection. Bow fronts tend to reflect EVERYTHING!
Here's a shot of my surface scum on the water. BOOM!! Just kidding, NONE!!
Here's an artsy-fartsy shot of my nymphoides aquatica pearling
Love the ideas, but rather costly to implement. (I'm guilty of worse, much worse...) Great thread! Two observations:
- It looks to me that the blend ratio is predicated on constant input pressure from both RO and tap lines. RO line is "wide open" and tap is controlled by a physical valve orifice size, right? So as your RO membrane ages and your RO output is reduced, your blend ratio will alter over time slightly, right? And in my case I'm on a well pump so my tap water pressure fluctuates with the bladder. Just some points to think about for others considering this setup.
- Your water change is not actually as much as you think, because the overflow water exiting the tank is not pure used water but a blend of new and used water.
Yepper on #1, the RO filter output will definitively slow down with time. I check for Chlorine breakthrough and hardness monthly or so, so far it has stayed fairly stable.
And yepper on #2. Below is a picture of input and output. I'm filling at about 175ml per minute, 45" away. So the new water is thoroughly mixed by the time it gets to the far side. ++shrug++ I'm ok with the fresh water loss because I avoid needing a drain timer and a drain solenoid valve.
This was posted for a different thread, but it's a detail of the overflow; The murkeyness is from installing a new canister filter. You're looking at a bacteria bloom here.
Same thread, detail view of the waste water draining via surface tension