DIY AUTOMATIC top-off

[Mihai]

Hi guys,

I'm well aware of top-offs (see http://reefs.org/phpBB2/viewtopic.php?t=56171) that have a big container that slowly keeps a rock-steady level in the sump.

The problem with those is that for those I need to fill in the top-off container itself (how lazy can one be? ). I have the output of my RO/DI filter somewhere very close to my sump, I just don't want to periodically top-off the container.

An easy solution would be to use a float valve on the output of the filter and use it to maintain the constant level in the sump. Unfortunately this will lead to a premature death of my RO membrane due to short cycling (turning on/off several times a day). What I'm looking for is a method to automatically refill the auto-topoff container only when it empties (or almost empties).

I thought about it and I don't think I can do that using only one float valve. Perhaps not even with two float valves. The only way I know how to do it is by using two float switches, a bit of logic (to implement a histerezis) and an electrovalve...

Question 1:
Do you have an easier idea? (that still allows me to be lazy ) ?

Question 2:
If the minimal solution is the one I mentioned (2 float switches, one relay, one electrovalve, 1 microcontroller) where the heck do I find a good electrovalve?

Thanks,
Mihai

 

[holry7778]

Could you do something very similar to the tank behind a toilet?

The Stimulus being when the auto-top-off being low using a weighted float, pulls the "level" and "flushes" the RO "toilet" tank into the topp off tank. This would inturn open the normal toilet float value to the RO filter and run until it fills the tank again...

No Electronics involved I don't think. If you try let me know hoe it works this is one fresh from the ol' knoggin. I'd be interested to see how well it works

Good Luck

 

[Mihai]

Nice knoggin you've got in there!
That looks really good - it satisfies all my long list of requests. I'll certainly post when I'll get it done.

But first, I'll have to figure out a way to make the bottom auto-top-off tank sealed such that the trick with the two tubes works.

Thanks,
Mihai

 

[holry7778]

Yeah That might not work then. sealed Hmmm ???

 

[Anonymous]

There's no reason the line going vertical has to be inside the top tank. It could be outside the top tank, and well above it, and still pull up the flush valve. That's a really good ideea.

 

[Mihai]

I was not talking about the top tank - I was talking about the bottom tank that should be sealed in order for the trick with the two tubes to work. In the mean time I realized that it's probably easier if I have the bottom tank drain into the sump through an IV drip line (shouldn't clog - it's RO water!) and through a float valve. The drip line is supposed to be insurance against a valve stuck "on".

Regards,
Mihai

 

[Fatal Morgana]

my suggestion:

water path: solenoid on timer-> RO->2.5 gal bladder tank->auto top solenoid-> sump.

The solenoid on timer will turn on 1 to 4 times a day, for a few hours, depends on your top off rate and the RO GPD rating. If your RO is 24 GPD (a gallon per hour), then you want to turn it on around 2.5 hour at anyone time. If your topoff rate is 5 gall per day, then you set the timer to be on for 2.5 hour at noon, and 2.5 hour at midnight, for example.

So the bladder tank will be full most of the time with 2.5 gal of water ready to dump. When the autotopoff controller sense low sump level, it will open the solenoid and drain the water from the bladder tank.

What you need:
Timer
NC solenoid for controlling input to the RO filter
bladder tank, 2.5 gal is most popular
auto top off controller, such as SpectraPure's LLC
2 hours to set it up.

 

[Mihai]

Negative. You're proposing to match the rate of evaporation without feedback not once but twice. It's impossible to match it once, let alone twice. You'll constantly have either spills of RO or empty containers, even without any malfunction. There is no way you'll make that to work over the long run - the evaporation rate is not constant.

What I'm trying now is to build enough redundancy in the system to avoid disasters in the case of a malfunction (either burn the return pump or drop my salinity).

Regards,
M.

my suggestion:

water path: solenoid on timer-> RO->2.5 gal bladder tank->auto top solenoid-> sump.

The solenoid on timer will turn on 1 to 4 times a day, for a few hours, depends on your top off rate and the RO GPD rating. If your RO is 24 GPD (a gallon per hour), then you want to turn it on around 2.5 hour at anyone time. If your topoff rate is 5 gall per day, then you set the timer to be on for 2.5 hour at noon, and 2.5 hour at midnight, for example.

So the bladder tank will be full most of the time with 2.5 gal of water ready to dump. When the autotopoff controller sense low sump level, it will open the solenoid and drain the water from the bladder tank.

What you need:
Timer
NC solenoid for controlling input to the RO filter
bladder tank, 2.5 gal is most popular
auto top off controller, such as SpectraPure's LLC
2 hours to set it up.

 

[Anonymous]

I'm confused--what do you mean by "the trick with the two tubes"? Are you talking about making the bottom container airtight to use as a Nurce type topoff?

The issue with float valves getting stuck 'on' is usually only a problem when dripping kalk. If you're just dripping RO water I think you'll be fine. The hinge won't even be in contact with salt water.

I'm really interested to see how this all works out.

 

[Mihai]

I'm confused--what do you mean by "the trick with the two tubes"? Are you talking about making the bottom container airtight to use as a Nurce type topoff?

This one:
http://home.att.net/~s-pawlowskis/topoff.htm

The issue with float valves getting stuck 'on' is usually only a problem when dripping kalk. If you're just dripping RO water I think you'll be fine. The hinge won't even be in contact with salt water.

Makes sense, but are you willing to bet your tank on that? I'd feel more comfortable with some redundancy. Ideally different technologies (float switch, float valve, drip, etc.).

Talking about an electrovalve - where can I get one?

Thanks,
M.

 

[holry7778]

Got it. An yet Another Idea stolen from the plumbing trade!

You'll just nee to add a sewer trap in there to maintain the vacuum. Now I know that there will be some testing required but hopefully it will work :?

 

[Mihai]

Got it. An yet Another Idea stolen from the plumbing trade!

You'll just nee to add a sewer trap in there to maintain the vacuum. Now I know that there will be some testing required but hopefully it will work :?

Man! That's pure genius! Either that or you're into designing plumbing contraptions for a living... The only thing that bothers me a bit with the design is that it doesn't have redundancy: if the vacum in the bottom topoff tank breaks I kill all my corals and inverts... I'll be thinking how to make it redundant...

One more thing: the vacuum might be broke by sucking the water from the u trap into the topoff tank. It might be a toss between stopping the drip into the sump and sucking the water from the u trap.

Thanks a lot for the time and ideas,
Mihai

 

[holry7778]

Man! That's pure genius! Either that or you're into designing plumbing contraptions for a living...

No I'm just a guy who like to tinker.

You'll have to raise the loop of the vcuum tub above that of the upper tank top off tank. I would think that the Vacuum pressure itself would not be able to over come the trap.

Who knows. Play with it and tell me how it works out. Should be pretty cheap to test too. That is why I try to think in plumbing equipment and not hum-fangled custom things.

Good luck!

 

[Fatal Morgana]

Negative. You're proposing to match the rate of evaporation without feedback not once but twice. It's impossible to match it once, let alone twice. You'll constantly have either spills of RO or empty containers, even without any malfunction. There is no way you'll make that to work over the long run - the evaporation rate is not constant.

What I'm trying now is to build enough redundancy in the system to avoid disasters in the case of a malfunction (either burn the return pump or drop my salinity).
...

Don't dismiss me so quickly
There is a feedback mechanism in my scheme. The autotop off is such mechanism, and there is no way you are going to have spills due to the topoff's malfunction. If after a few days, and the bladder getting less and less full, you can add additional time and/or add additional session to the solenoid to be on so that the RO can fill the bladder tank to capacity.

What matter is the max. amount of RO water you tank need per day. By having a smaller tolerance to the max. amount and actual amount, you will limit how much water can spill if your return pump fails. for example.

 

[Anonymous]

I'm confused--what do you mean by "the trick with the two tubes"? Are you talking about making the bottom container airtight to use as a Nurce type topoff?

This one:
http://home.att.net/~s-pawlowskis/topoff.htm

The issue with float valves getting stuck 'on' is usually only a problem when dripping kalk. If you're just dripping RO water I think you'll be fine. The hinge won't even be in contact with salt water.

Makes sense, but are you willing to bet your tank on that? I'd feel more comfortable with some redundancy. Ideally different technologies (float switch, float valve, drip, etc.).

Talking about an electrovalve - where can I get one?

Thanks,
M.

I'd be willing to bet my tank on a float valve delivering RO water before I'd bet it on the Nurce type system. Nice and cheap, but prone to failures IME until you tinker with it a bit. For the level of automation you're asking for here it would be 1000 times simpler to just buy a dosing pump, put it on a timer, and have it dose the entire day's topoff over an hour or two. All you'd have to do is periodically ramp up or down the rate, which is just pushing a button VS. making sure this Rube Goldberg plumbing device is still working and all the tanks are filled up.

Just me 2 cents...

 

[Fatal Morgana]

FWIW, I agree with Matt. However, dosing pump is a more expensive alternative than my suggestion above. Another 2 cents.

 

[Mihai]

FWIW, I agree with Matt. However, dosing pump is a more expensive alternative than my suggestion above. Another 2 cents.

Not only that - he's also betting on an open loop system - matching the evaporation rate with the delivery rate of the timed pump. For that there are simpler systems: just use a drip line and as long as you're doing RO and not kalk you should be fine.

I have to agree with the assesment on Nurce's system reliability: unless I find a system to guarantee that I'm not losing presure in the nurce system (e.g., lots of silicone) I'm not betting on it. If it's guaranteed not to lose preasure it looks pretty sweet as it has no moving parts to get stuck (either open or close).

M.

 

[Mihai]

Got it. An yet Another Idea stolen from the plumbing trade!

You'll just nee to add a sewer trap in there to maintain the vacuum. Now I know that there will be some testing required but hopefully it will work :?

It looks more and more unlikely that it will work: as long as you have water in the middle tank that water will want to go into the bottom tank and through it in the sump - when it manages to do that it will flush the top tank again: the RO output practiacally flows into the sump at all times - what keeps the water from the second tank from going in the Nurce device? I'm yet to find a solution to fill the nurce device when empty while maintaining vacuum when it's not...

M

 

[Anonymous]

FWIW, I agree with Matt. However, dosing pump is a more expensive alternative than my suggestion above. Another 2 cents.

Not only that - he's also betting on an open loop system - matching the evaporation rate with the delivery rate of the timed pump. For that there are simpler systems: just use a drip line and as long as you're doing RO and not kalk you should be fine.

I have to agree with the assesment on Nurce's system reliability: unless I find a system to guarantee that I'm not losing presure in the nurce system (e.g., lots of silicone) I'm not betting on it. If it's guaranteed not to lose preasure it looks pretty sweet as it has no moving parts to get stuck (either open or close).

M.

I guess I live in a climate where the evap rate is pretty constant. I usually set it and maybe update it like once a month. With the high end digital ones you can calibrate it down to the centiliter.

 

[Fatal Morgana]

solenoid valve has much better reliability than any mechanical flopper with chain, lever and siphon trap, and I strongly suggest you consider other alternatives. As usually, IMHO.