DIY Auto-Top-Off (ATO) System

[JWeb]

Gang,

Several people have requested that I do a write-up for my DIY ATO system. Unfortunately, I didn't take many photos during the actual construction of the system, so I don't have step-by-step images. But, I do have photos of the finished product and will be happy to fill in any of the details if anyone decides to put one together.

As we know, there are many ways to facilitate an ATO system - and each seem to have many pros and cons. When I set out to build my ATO, I had the following objectives in mind:

1. It had to be automatic - requiring virtually no interaction on my part.
B. It had to be reliable - since I don't like replacing burned-up pumps when the sump runs dry or using my wet/dry vac to suck up a flood.
3. It had to be invisible - my reef tank is in my living room and I would like it to blend seamlessly into the room (all of my DIY projects have been significantly focused on this goal.)
FOUR. It had to keep the reef happy (isn't that the ultimate goal?)

I have a 100GPD RO/DI system that I use for drinking (RO) and reefing (DI). The pressure tank associated with the RO/DI system stores RO water on tap at the sink. The DI water is generated from this supply as it passes through the DI resin. Thus, the first order of buisness was to create a DI reservoir.

I ran 1/4" tubing from the RO/DI unit to a 7-gallon bucket housed in a cabinet beside the reef tank (my reef is only 30 gal display + 15 gal sump, 7 gallons is plenty for my ATO). I installed a USplastics.com float valve in the side of the 7 gal bucket by simply drilling a hole of the prescribed size. See image below:

http://www.jeffreywebster.net/IMG_0995.JPG

As you will see, the top-off water is fed to the sump via a powerhead. You can the powerhead installed in the bottom of the bucket. The plastic tubing is obviously to direct the pump's output to the sump (be sure you don't put the end of the line below the surface of the water in the sump - you don't want to create a syphon when the pump turns off after topping off!)

Thus, I have a constant 7-gallon supply of DI water. The RO/DI system does not run constantly since it fills the bucket to the level determined by the float valve and the flow ceases.

Next, I installed two float switches into the sump (in the return section - where one sees fluctuations in level due to evaporation). The float switches were obtained on eBay for approximately 2$ each. The lower switch is the "low-level" switch that will activate the ATO system. The upper switch is the "high-level" switch - which will be used to signal the ATO system that the sump is full. Of course, placement of the float switches will be determined by the acceptable "low-level" and "high-level" for your sump. I ran the leads from each of the float switches to the area where I would install the control box - the "brain" of the ATO system.

The "brain" of the ATO system is a latching relay. I selected a 5A, 120VAC dual-12VDC-coil, DPST latching relay (digikey.com part no. 255-1056-ND, 7$). This relay will easily handle the amperage draw from a small powerhead (there are cheaper relays that will handle less current - select appropriately). The dual-coil arrangment allows for "set" and "reset" signals - which, as you will see, are provided by the float switches in the sump. I chose 12VDC coils since that is the DC power supply I use to power other items in my reef control system. However, if you aren't currently using DC power, you can grab any old "wall wart" type DC power supply and select the relay coil voltage to match what you can get your hands on. The DPST switching arrangment is necessary if you have a GFCI installed on the circuit to power the pump (you must switch both the hot and neutral AC lines, as I will explain).

To complete my system, I obtained:

1. A standard 120VAC outlet (2$, Home Depot)
2. Miscellaneous hardware (screws, nuts, nylon spacers - 3$, Home Depot)
3. A plastic project box (6$, Radio Shack)
4. A project board (2$, Radio Shack)
5. IC socket (1$, Radio Shack)
6. 120VAC three-prong plug/wire (2$, Radio Shack)

You will also need wire, solder and a soldering iron.

The rest of the construction is wiring the circuit and fitting the above to the project box.

The circuit is relatively simple:

The low-level float switch is wired to the "set" coil of the latching relay. (See DigiKey datasheet (http://rocky.digikey.com/WebLib/Aromat/Web%20Data/DSP%20Relays.pdf))

The high-level float switch is wired to the "reset" coil of the latching relay.

Essentially, the float switches complete a 12VDC circuit (DC+ -> float switch -> relay coil+ -> relay coil- -> DC-).

The hot and neutral leads feeding the AC outlet are soldered to the normally open pins of the latching relay.

The hot and neutral leads from the 120VAC three-prong plug are wired to the supply pins of the relay. (Again, see DigiKey datasheet).

It is important to switch the hot and neutral AC lines at the same time if you are using GFCI - otherwise, the GFCI will trip everytime the relay switches.

Thus, when the "set" coil is energized by the low-level float switch, the relay latches and closes the AC circuit (energizing the AC outlet). The outlet will remain powered until the "reset" coil is energized by the high-level float switch. Viola - ATO.

With the wiring complete, as shown in the photo below, I used nylon spacers to mount the project board and AC outlet on a small piece of acylic - which was then fastened inside the project box:

http://www.jeffreywebster.net/IMG_0993.JPG

I then cut openings for the AC outlet in the aluminum cover for the project box - makes everything nice and neat:

http://www.jeffreywebster.net/IMG_0998.JPG

Like anything else, there are "failure modes". In my system, the two components that I worry most about are the float valve in the 7-gallon bucket and the high-level float switch. If either get "stuck", you will have a flood. You could always install redundant high-level switches in the sump - it adds to the wiring confusion, but will virtually eliminate the potential for a stuck high-level switch. Like anything else, I think proper maintenance will help to prevent any mishaps.

I hope this was useful (and complete!). I'm here to help if I can.

Jeff

PS> Please excuse any blatant spelling or grammatical errors!

[NateHanson]

Nice work Jeff. The DIY forum is coming along nicely here at BRS!

[nunofs]

Jeff, thanks for sharing this, it looks awesome! I agree that some level of redundancy would be good to have, from what I read the float switches/valves tend to get stuck...

Thus, I have a constant 7-gallon supply of DI water. The RO/DI system does not run constantly since it fills the bucket to the level determined by the float valve and the flow ceases.


This is the only thing in your project that I have questions about... with a single float valve in the bucket, I'd think that the RODI would be running very frequently, everytime the water level in that bucket drops a little bit... I know you have a RO reservoir, but if it works anything like mine does, as soon as there's a bit of demand from that RO reservoir, the internal pressure drops and triggers the solenoid that starts up the RODI filter. So I'd think that you'd be cycling the RODI on and off very frequently, which wastes quite a bit of water.

I believe that to totally eliminate this problem you'd need to setup a double valve inside the bucket, much like what you did for the sump: this way the RODI would provide water when the level in the bucket dropped below a certain level and would shut off when the water reached the top float valve, being inactive at all other times and achieving better efficiency.

Nuno

[NateHanson]

Nuno, Water only leaves his bucket (and thus the RODI is only activated when the sump relay is switched on, energizing the powerhead. The frequency of that will be determined by the spacing of your floats in the sump.

The improvement I would make, Jeff, would be to put your latching system in the bucket reservoir, (connected to a solenoid valve on the RODI, instead of to a powerhead), and have the sump constantly topped off from the bucket by a mechanical float valve in the sump (just swap what you have in the sump with what you have in the bucket). That way the sump will be topped off constantly, and salinity will be rock solid. The only place the water level will vary is in the bucket. The way you have it now, the sump is being topped off in quanta, rather than continuously, so salinity will go up and down (although the differences will admittedly be very very small). Also, if you decide to add a kalk reactor at somepoint, you wouldn't want the kalk reactor fed by the powerhead in bursts, because it would stir up the solid kalk just as it was being added. Also it would be best not to add kalk in big bursts, but rather a constant trickle. Again, that's only an issue if you decide to add a kalk reactor later (an easy DIY! - I'll post the pictures soon). This modification will also allow you to decrease the cycles the RODI experiences, because you can space your upper and lower float switches farther apart in the bucket, giving more time between refills. A final advantage of this modification is it gives you an easy opportunity for a redundant fail-safe mechanism. Just have another float valve where the RODI water enters the bucket, set above the upper float switch, so if your electrical system failed open some how, the mechanical valve would prevent a flood, and your ATO would continue to operate until you fixed the electrical anti-cycling system (the only difference would be that your RODI would cycle frequently).

Very cool project. I definitely don't have the electronics know-how to come up with this, so I'm glad you described it for us.

Nate

[nunofs]

Nate, I think we were saying pretty much the same thing, maybe I didn't explain myself well...

I was also suggesting, like you did, that Jeff should add to the bucket a similar setup to what he has in the sump, with two valves, exactly to limit the cycling of the RODI.

What I hadn't thought of was switching them around and using the float switch for the sump, to provide a more constant water level, but that makes sense!

Nuno

[NateHanson]

Yeah, the way I see it, the only reason not to just use float valves is for RO unit cycling, so I don't think a second float switch arrangement in the sump would add anything, if you had the double float switch in the reservoir.

[nunofs]

You're right, it wouldn't serve much of a purpose... except if it was configured as primary and backup switches in series (as opposed to low/high level switches), to provide redundancy. That's something that float valves alone don't really address.

Nuno

[NateHanson]

Yep. Very true. I've been hoping that I could find an in-line float valve somewhere. That way you could install it higher than the terminal float valve, and it would act as a high limit safety valve. But so far it doesn't seem to exist. :(

Nate

[JWeb]

Great suggestions, guys. I will definitely consider changing the arrangment. Since I have so little reef experience (actually, none, other than building this stuff), I never really considered the swing in salinity. But, as you suggest Nate, the total top-off volume is rather small compared to the whole volume of the system. I just got my refractometer - so it will be interesting to see what the changes are. (The swing-arm
units really do suck.)

nonofs, my RO/DI doesn't run continuously - but I also don't have MH lamps over the tank, yet - just PC. The evaporation rate has been about 1 gallon / day (my low level in the sump is about -2 gallons). So, the RO/DI has only been coming on every other day to fill the bucket. I wondered how the evaporation rate might change with MH lamps... and feared it might be substantial. The ideas you and Nate have suggested would certainly help (although, I am limited to about a 5 gallon swing in the bucket... so I might just cut RO/DI cycling in half). I suppose I would need a 55 gallon drum to really reap the benefits. But, that violates Objective 3, above!

I will probably order a solenoid valve, anyway - to put an added layer of safety into the system. The fun never stops!

Jeff

[NateHanson]

If you're getting 2 gallons between your high and low, I'd say that's pretty good as far as the RODI cycles go (it's probably cycling more often when you use it to get a drink from the RO end). 2 gallons out of 90 is about 2%, so you're getting a salinity swing from 0.025 to 0.0255. Not that big a deal (but still, a variation that could be avoided). To me the real motivation for rearranging your system to have trickle top-off in the sump is that you can use a kalk reactor, and that will definitely come in handy when you get into SPS, if the interest of JWeb objective #1.

With MH your evaporation rate might go up another gallon/day. Depends on how much MH you use, and how much you have to use fans and surface turbulence to dissipate heat. I lost about a gallon a day on my 30 with one 175 MH and 2 VHO tubes, and frankly I don't know what I'm losing now with my 65g (total 80g) with 3x175w MH, because I've had a float valve hooked up to the RODI.

Nate

[JWeb]

Nate, I remember now why I use a powerhead as opposed to the solenoid valve arrangement: my DI reservoir (the 7-gal bucket) is at the same level (floor) as the sump. Therefore, I have to pump from the reservoir to the sump. With your idea of a solenoid valve controling the flow of DI from the RO/DI unit into the reservoir (switched by floats in the reservoir itself) and a float valve in the sump, I would need to have gravity feed to the sump (unless I am missing something in your descriptions).

I will probably want to build a kalk reactor at some point, but I simply don't have room in the stand or the cabinets (book case) near the tank. What I believe will happen shortly is that I will want a larger tank sooner rather than later (isn't that how it always goes?) and will plan to build a better, more accommodating stand for the sump and ancillary equipment. But, I should probably get this tank set-up before I start drawing plans for a bigger one!!

Anyway, thanks for the tips guys - very valuable. Your improvements defintely address the weaknesses and "reef-ability" of my ATO system. So, I'll call the above: ATO v1.0.

Jeff

[NateHanson]

Yes, that's a problem with my suggested mod.

And by all means, get that tank going! Put some livestock in it! Grow some corals! Isn't that the point of all this? :)

[heavydc2]

At the first club function I ever went too we made auto top off float switches. I think your idea is great, this is just another method along the same lines you did here.
We bought the float switches (ebay or supply place), and simply put them on a big plastic hook (like the ones plumbers use to hold pipes to the wall) and some pvc caps, pipe and plastic bolts. Once assembled exactly like you have it, one higher then the other for redundancy. Then wired them into a external vacuum pump http://www.marinedepot.com/md_viewItem.asp?idproduct=KTALP with a good amount of electrical tape to cover the exposed wire, it was a very simple and cheap project. I think the total cost was about 20 bucks with pump and all. I can elaborate if you would like, I have improved on the hanging system on mine, but thats only because I have a little acrylic to play with.

[RobboT]

I built one last night like Dave was talking about with float switches. I used 1/2" PVC to hold them which is nice because the wires run inside the PVC. I'll take pics when I build the next one (building for a friend) because it'll be prettier. Probably cost me about $30 all told not including the pump.

[JWeb]

heavydc and robboT, are you using the float switches to switch the 120VAC? Also, are you using a latching relay (if not, how do you establish a low vs. high signal?). I'm curious about how the switches look mounted inside PVC. Please share a photo.

Jeff

[nunofs]

Jeff, take a look at http://autotopoff.com ... the kits he sells also have the switches mounted inside PVC... it protects the switches both from being hit by other equipment and also from "waves"/splashes in the water (that could trigger the switches unnecessarily).

Nuno

[RobboT]

Jeff -

Yes the switches are rated for 120. The switches are wired in series and both set in normally closed position. One is set at the level that I want my water, the other about 6" above that. When the water level drops the bottom switch closes and turns on the pump because the top switch is also closed. When the watr level rises the bottom switch opens and shuts off the pump. If by chance the water level rises and the bottom switch stays closed the water will then open the top switch and shut off the pump. Works great. Just have to make sure to keep the switches free of salt. I have a diagram. I will JPG it and post.

[NateHanson]

But how much water comes in before the float switch shuts off. I would think it would be prone to very frequent cycles, since it's shut-off level is the same as it's turn-on level.

[RobboT]

Here it is. I just added the wiring part. Im not an electrical engineer so its really bad but I think you get the idea. The switches are $11 each. I used an outdoor outlet box with the spring loaded covers which I think cost me $6 or so. The outlet was like $1.50. The power cord was probably the most expensive part technically because I used an industrial rated grounded power cord but it was one I had lying around that wasnt being used cuz it was only 6 feet long so no big loss anyway. You could probably get an adequate cord at HD for $5 or so. So ~$35 or so. I sealed the extra knockouts in the outlet box with caulking. Im using one of those Aqua Lifter pumps which I think cost me $15 at Aqua Addicts. It feeds from my home made kalk dripper which cost me $10 to make. So all told Im around $60 into this thing.

Been thinking about making them and selling them actually. Just a little extra on top maybe to help fund the "gallon a day tank" set up.

http://home.comcast.net/~robbo_t/ReFloatSwitchModel2.jpg

[RobboT]

BTW sorry for the text on the drawing. I did it that way so I can send it to two friends for info on constructing theirs.

[RobboT]

But how much water comes in before the float switch shuts off. I would think it would be prone to very frequent cycles, since it's shut-off level is the same as it's turn-on level.

The switches dont have a hair trigger on them so it works out fine. I also discharge in my fuge not in my sump so theres even more lag time before it shuts off. I'd say the fluctuation in the sump (a 10g tank) is 1/2" to 3/4" which means it doesnt pump until theres about 1/2 gallon required. I've been using 1-2 gallons a day lately so that means the pump only runs 2 - 4 times a day.

[JWeb]

Drawing is great, RobboT. As Nate suggests, however, it seems that you are really only switching the pump on the low-level switch... and the pump is probably cycling quite often. Is that true? Also, do you worry about having 120VAC wired directly into the water? Although the wires probably provide great insulation, the risk is much higher, IMO.

My system only wires 12VDC into the water and uses two switches - LOW=ON, HIGH=OFF. In addition, the current flowing through the switches is only what is necessary to pull the relay coil (a few milli-amps). I suppose, though, that even 12VDC at low current would cause issues if current leaked into the reef.

So, as I said in the very beginning, there are many ways to accomplish this!

Nunofs - I've looked at autotopoff.com several times and only now just realized I was looking at PVC caps covering the switches. Thanks!

Jeff

[JWeb]

HA! I always forget to read the next page of posts to find the answers to my questions.

Jeff

[RobboT]

Also Jeff, the switches are sealed so there is no exposed wiring in the water and its in my sump so there is no chance for voltage in the reef itself.

Without using a complex setup of float switches you really always have to rely on a single switch. More importantly this kicks the $--- out of the "one-drop-per-second-hope-the line-doesnt-clog-hope-one-per-second-isnt-too-fast" method I was using before :D

[NateHanson]

Also Jeff, the switches are sealed so there is no exposed wiring in the water and its in my sump so there is no chance for voltage in the reef itself.


Well, its a remote possibility to have an electrical short in the sump, but actually you're whole SW setup is electrically contiguous. As long as there's a solid channel of water from the sump to the display (return line or drain will work for this) the whole setup should be electrically live if there's an AC leak in the sump.

[RobboT]

Very unlikely that any significant voltage could make that trip. If this really was a concern then there is no float switch scheme to avoid it. The sensors themselves are just as likely to cause the voltage issue as the wires in the scenario your talking of.

[RobboT]

Here are some very bad pics of my switch setup. One of these days Im going to learn how to get the camera to focus properly.

Picture 1 is the switch rig. Picture two is the box that you plug your pump into. The orange cord coming out of the back of that box goes into the wall outlet.

It ain't pretty, I know. I literally made it from leftover pieces of PVC that I had lying around. The next two are going to be much prettier.

[NateHanson]

Yeah, I agree there's no reasonable chance for this to happen, but it's all one bucket of electrolyte. I'm really not saying anything negative about your switch arrangement. Sounds good to me. I was just making a note about the electrical continuity of the sump and display.

[JWeb]

RobboT, it says right there that you are advocating against whizzing on the electric fence! ;)

I like how you hide the wires in the PVC pipe - that's something I need to do. What about putting PVC caps over the switches as I've seen on autotopoff.com? Any reason you might have chosen not to do that?

If we put all of these ideas together, we'll have one hell of an ATO system.

Jeff

[RobboT]

Hmmm somehow a post I made got lost, weird.

Nate - I know you weren't. :) Unfortunately reefkeeping means lots of salt water near lots and lots of electricity. Nature of the beast I guess. :)

Jeff - the onyl reason is that I didnt have any lying around. I've had those switches in my basement for over a month. Being the procrastinator that I am I never got around to figuring out the parts I needed and actually building one. This guy that I am mentoring needs a top-off by early Dec so he can go home with a little peace-of-mind. So finally Tuesday night I went to the basement, dug out a pile of 1/2 PVC and put one together. For the next two Im building I will go to HD and get the right parts. :)

[NateHanson]

This is going to be a great article for the newsletter. At some point I'm going to impose on each of you to write a short blurb of your ATO solution, and I'll put them together into an article with "four solutions to automated top-off" or something like that. So take some pictures now while it's easy to do so.

Thanks, Nate

[RobboT]

Nate if you want you can refer back to "A Tale of Two Drippers" which I wrote for an old newsletter. You might be able to use some info in there in preparing a new story.

[NateHanson]

thanks Rob. I'll look that up.

[JWeb]

Is my write-up here too comprehensive, Nate? Are you looking for a summary?

Jeff

[RobboT]

We did an in-place setup for Dawn (skyedolphan) this past Wed that was my favorite so far. Forgot to take pics. I used horizontal switches instead of verticals, drilled them right into the wall of her 100g stock tank. Wired up and set the pump in her RO/DI reservoir. Very neat and tidy. Going to build another hanging version with vertical switches tomorrow, that one should be neater than the frankenstein's monster I put in my tank so I'll get pics.

[JWeb]

Sounds cool! What are these horizontal switches like? Any advantage over the vertical switches? Did you use a latching relay setup or your "hair-trigger" configuration? ;)

Jeff

[RobboT]

Hair-trigger!! I'm no electrician so I do what I can.

The switches look like the pic I attached.
www.mcmaster.com page 457 of the online catalog.

Being a shortsighted cheapskate I saved $2 and got the inside thread ones. It worked out because of the setup Dawn had. I just drilled my 5/8" holes in the side of the sump, threaded them and screwed the switches in. I like the horizontal switch better than the vertical. In the future I will order switches that are threaded on the outside as well so I can just screw it into a fitting for a hang-on type of setup. If someone wants a direct drilled sump setup (as I think Lam does ;)) I will get those because it makes a very neat and clean setup.

[JWeb]

I like those... as you say, they will easily integrate into a PVC support structure.

I'd love to see some pictures of your creations!

Jeff

[RobboT]

I will get some but dont hold your breath. :)

[RobboT]

OK I dont have any of the horizontal install but I have one of the vertical switches on a hang on rig that is much prettier than the ones above.

The two switches
http://home.comcast.net/~robbo_t/Resize2ndFloatSwitch05.JPG

Wider view
http://home.comcast.net/~robbo_t/Resize2ndFloatSwitch06.JPG

Installed in sump
http://home.comcast.net/~robbo_t/Resize2ndFloatSwitch08.JPG

The box you plug your pump into
http://home.comcast.net/~robbo_t/Resize2ndFloatSwitch10.JPG

The person I made this for IM'd me this morning to tell me he's going through manual top-off withdrawal :) He's constantly checking the tank only to find out that its fine.

For a small fee I will make one for anybody who wants one.

[JWeb]

That looks awesome, Robbo. Thanks for the pictures...

Jeff

[ceige]

hmmm whats a small fee? that thing is great!

[marco67]

Nice work guys,
Just a word on switch cycling / reliability. I'm using the same switch (vertical) to switch on a DIY I did years ago. Mine switches a relay that in turn switches a peralistaltic pump. The "throw" on these is close to an inch so there is no hair trigger effect.
BTW
My single float switch version is running now for over 3.5 years without drama.

[RobboT]

hmmm whats a small fee? that thing is great!

I was thinking of asking $50 for the double switch setup. It costs about $40 in parts, a little less if I plan ahead and don't have to buy things last minute at Home Depot.

[RobboT]

On this subject, I am not an electronics expert so maybe someone can help with this. A latching relay is what is needed to have the pump turn of on the bottom float and off at the top float. When the switch closes on one side of the latching relay the power goes on, when the switch closes on the other side of the relay the power goes off. Everywhere I look these things are $30. Can anyone explain to me what the difference between a plain old relay and a latching relay is?

[NateHanson]

I believe the purpose of a plain old relay in these setups is just to take the 12V signal from a sensor (the switch in your sump) and close the 120V circuit for your powerhead. It avoids having 120V in the sump.

A latching relay does just as you say, and adds a separate on and off signal circuit for the relay.

All that said, I'm a furnituremaker, so you'd be wise to take it all with a grain of salt. :)

[nunofs]

A latching relay basically "keeps state", ie, maintains its current setting, until it receives a new signal at its input... it's like a mechanical flip-flop circuit. On the other hand, a normal relay is only active while there is a signal at its input.

Nuno

[clamm]

i thought that a latching relay stays on "latched" until its hit with another signal. like you press a button once and the light stays on till you hit it again.

[NateHanson]

Sounds like the same thing as Nuno said Casey.

[clamm]

yup...same thing...posted at the same time almost :)

[NateHanson]

yep, sorry.

[RobboT]

OK then why go to a relay at all? Why not just wire direct?

[nunofs]

Because the switches are not wired exactly as you described above: the bottom switch turns on when the water level drops, triggering and latching the relay... it turns off when the water level goes above it, doesn't matter to the relay, as it's already latched on the active state; then when the water level reaches the top switch, it turns on and trigers the relay, which goes to inactive state. As evaporation makes the water level drop, the top switch eventually goes back to off (no impact on the relay), and the bottom one eventually turns on again, repeating the cycle.

Anyway, long explanation, hope it makes sense, but the important thing is that the two switches are wired differently, the bottom one is inverted.

Nuno

[nunofs]

Btw, this doesn't apply in case you're using the second switch as a redundant backup... in that case they're both wired the same way... and you don't need a latching relay in that case because the switches will be on for as long as the pump needs to run so the relay doesn't need to keep any state.

Nuno

[RobboT]

Sorry nuno I should have said "in the case of the standard relay". I understand the latching relay concept. Of course then I read back and saw the post by Nate that I missed about using a 12V source to control a 120V circuit and now it makes sense.

So . . . anyone know where to get latching relays inexpensively?

I should say latching relays that are entirely AC based.

[Tstone]

Hey Rob I want one of your ATO gizmos. Very slick. Makes mine look very crude.

[dedfish]

Here's a link to my auto-topoff. I haven't really used it full time yet, but all the test's I've done have worked out great. I just finished re-engineering the float switch bracket so it's a little different than the pics. The float switches have to be in the tank in this setup, or it would have been much easier. I use a relay so I don't have 120v flowing through float switches sitting in saltwater. There is a diagram at the bottom.

http://216.235.242.50/forums/showthread.php?t=2318&highlight=auto+topoff

[RobboT]

OK Marvin, you got it. Looks like Im going to be building a few of these. :D

As far as the 120V in the water issue . . . I tested my tank last night and have no stray voltage in the sump so Im not so concerned about it. Besides I've got a powerhead hanging directly in the main tank, and a 500W heater, a Mag9.5 and 2 Sedra pumps on my skimmer in the sump. :)

[Tstone]

Thanks Rob!
I am with you on the stray voltage. I have a bunch of 120V in the water. :)

[dedfish]

Rob, no stray voltage today...but what about tomorrow or the day after that? What about the day you have to move something or disturb the wires, or god forbid spill some water or have a leak. All these items you buy have been UL certified and approved. A topoff system you build which uses float switches NOT designed for that much voltage is a hazard. If you want to run 120v through a very fine guage wire (all the float switches I've seen have very thin wire) and then stick that in a tank off saltwater...very well then. Just check the specs of the switch and make sure it can handle it, my bet is the switch can't.

I just think that $20 in parts is worth my life and yours.

[JWeb]

Rob, also keep in mind that the float switches have a maximum current rating - which is consumed by the wattage/voltage combinations of all of the equipment running off of that switch. If I remember correctly, most of the float switches I was looking at were rated for 0.7A or so - which is approximately 80 watts at 120V (conservatively). It doesn't take much to use up all of that power. If you exceed the current rating of the switch, it could melt or (possibly worse) have the contacts welded shut - making it no longer a switch (= FLOOD).

(I also agree that stray voltage is still a concern... and although you have heaters and pumps in the sump, they are well designed to handle and isolate the 120V... the float switches are not, IMO.)

Also, I believe that you were looking for a source for latching relays that weren't 30$. If you review my original post in this thread, you will find a link to the latching relay I used for my ATO. I got it from DigiKey.COM and it was cheap - I think 5 or 7$.

Hope this helps...

Jeff.

[RobboT]

thanks Jeff, this thread is so long now I didnt even recall that relay link. :)

[RobboT]

OK we can talk on this 'til Im blue in the face and since I've become known for doing that . . . why the hell not.

12V is more than enough to kill you. I dont think that $20 makes me any safer.

The switches I'm using are UL approved. They are rated for 0.28 amps at 120V AC. I'm not sure what you know about UL but it only applies to certain uses of certain things, i.e. these switches are actually designed for use in HVAC Condensate systems so the UL approval will most likely only be valid in that application.

I would agree that in reality there is no amount of money worth my life however we take chances every day with a million little things, nothing is risk free. Just by getting out of bed I am risking death. Just because something is UL listed doesnt mean it wont fail, just because you've used less electricity doesnt mean it wont kill you. I'm confident that the system in my tank is as safe as I can make it within my own limits of risk. I'm confident that if something goes wrong with this system that there will be indicators to that failure that will be obvious to me before I put my hand in any current filled water that would kill me.

Now on to experience, I have had voltage in my system on two occasions. Once due to a 250W heater another due to a powerhead. In the case of the PH the current was directly in the main tank, there were no ill effects on my tank inhabitants only on me when I poked my hand in there. The heater was in the sump and I only found out about the problem when I stuck my hand in the sump to pull my return pump for maintenance. It had probably been that way for some time and I had my hands in the main tank prior to that and felt no stray voltage in the tank as a result.

So short of removing all the electrical devices from my system I cannot eliminate the hazard. Awareness is the key.

[Tstone]

Smiley face impression of Rob.

http://smilies.sofrayt.com/%5E/_950/cleverman.gif

[Tstone]

Oh yeah I agree Rob.

I took a chance and got out of bed this morning and lived to tell about it. http://smilies.sofrayt.com/%5E/a/wavey.gif

[NateHanson]

I'm not trying to get into something here :), but 12V actually can't even shock you, let alone kill you. The potential in 12 volts is barely high enough to overcome the resistance in your body, so 12 volts generally won't be able to shock you if you're holding pos and neg 12 volt leads in opposite hands. 120 V on the other hand has more than enough potential to travel through your body.

[RobboT]

I give up :) Heading out to connect my car battery to my testicles to prove Nate's theory :D

[smcnally]

I'm not trying to get into something here :), but 12V actually can't even shock you, let alone kill you. The potential in 12 volts is barely high enough to overcome the resistance in your body, so 12 volts generally won't be able to shock you if you're holding pos and neg 12 volt leads in opposite hands. 120 V on the other hand has more than enough potential to travel through your body.
Not to dredge up an old thread, but I just read this and was wondering if this is what a lot of people think. 12Volts can kill you just as well as 1 Volt can kill you if it has the right amount of amperage. One thing I was taught in electrical training was this quote "It isn't voltage that kills you, but the amperage in it" Did you know that when you get a static shock on your finger tip when touching a door knob etc., that that is a minimum of 3,000 volts. Crazy, huh?