DIY Auto-Top off with Solenoid - Page 5

NoFear29 · #**101** 09/01/2003, 01:50 PM

Zeph.... I am glad you are still keeping up with this thread. I have you Auto-Topoff running for some time now, no problems. Thanks for the great DIY project.

I just completed my DIY kalk reactor, and hooked it in with the auto-topoff. I had a leak last night because my Martha Stewart lid popped off (Martha Stewart Cap Link) and caused a leak. It seems that my water pressure might have been a little to high for the lid. I have since added some teflon tape and a little weldon 16 to the inner surface to create some more friction. It seems to be holding.

Zeph, I just wanted to make sure I ran this thing correctly with the DIY topoff.

I have it plumbed this way. RO/DI unit to solenoid, solenoid to kalk reactor, kalk reactor to secondary (fail safe) float switch in the sump. Does that sound right to you? I was a little leary about running the Kalk through the solenoid. How do you have yours running?

NF29

NoFear29 · #**102** 09/01/2003, 07:27 PM

Update:

The Martha lid popped again causing another small leak. I am thinking about cuttint the whole top of the reactor off and installing some kind of test cap or maybe the 4" to 2" reducer with a 2" union.

Also discussing this on another link:

Other Link.

Steven M · #**103** 09/01/2003, 07:46 PM

Quote:

Originally posted by NoFear29
Update:

The Martha lid popped again causing another small leak. I am thinking about cuttint the whole top of the reactor off and installing some kind of test cap or maybe the 4" to 2" reducer with a 2" union.

:<

I have a 4" test cap no my reactor. The 4" test cap well not fit a 4" pvc fitting, it well fit a stander 4" pvc pipe.

Zephrant · #**104** 09/01/2003, 09:39 PM

NF- When you run it though the kalk reactor, you have to remove the float valve. Even the big fancy kalk reactors cannot handle the 40-60 PSI that the RO/DI unit can put out. When the float goes shut, the RO/DI builds pressure in your kalk reactor until it pops. The more you seal up the unit with tape/glue, the more pressure builds until it pops again.

I ran some tests on a 6" cast acrylic reactor. At 5 PSI, the 3/8" top flange had 165 pounds of pressure on it, and it bowed over 1/8". By going to 8 stainless steel bolts, and torqeing them down tight, I could get to 10 PSI, but it was starting to look like a balloon. With 331 pounds of pressure on that lid, if it gave way, it could cause some serious damage.

I pressure test my units to 5 PSI now, and warn that they can't take normal house water pressure. Luckily, they start to leak past the O-ring when they get too much pressure in them.

The best solution is to install another float switch inline with the first one, mount it 1" higher, and flip it upside down so it is normally closed, and opens when the water gets too high. Then you have to have both switches fail before you have a problem.

HTH

Zeph

NoFear29 · #**105** 09/01/2003, 10:05 PM

Zeph, thanks for the info. Should I order the same float switch again? How do I wire it inline.. Is it a matter of just soldering it in the same way the other one was? 1 to AC cord and the other to the solenoid?

My brain is somewhat fried from messing with this Kalk reactor tonight so your help would be, and always is appreciated.

mattboy · #**106** 09/02/2003, 12:34 AM

Zeph; I suppose the "one way" aspect of the solenoid makes sense. When I called McMaster-Carr, they told me to call back on Tuesday (long weekend) to speak with some in the tech department about these solenoids. They're going to send me another one. I'm thinking of using both; one before the R.O. filter, one after. I'm assuming the pressure on the output side of the filter is quite low; right? Can I just wire them both up to the same 2 wires coming from the float switch/A.C. plug? i want them to be activated at the same time.

That way, I'd have a "fail-safe" system. Both solenoids would have to fail open or leak to allow water to continously pass to the sump, and since the whole thing is on a timer, the float switch and the timer would have to crap out to allow continuous power to the solenoids. Any single component; timer, float switch, solenoid, could fail and I'd still avoid a flood. Of course, they ALL have to work to keep the sump from running dry eventually, but even if that happened, my closed loop would keep the tank alive for days.

BTW, getting this thing plumbed was a huge PITA. I had to crawl under the house (1 foot crawl-space) find the right spot to drill up into the wall, fish the hose and wires through, feed them through a conduit so they're protected under the house, etc...but if it works trouble free it'll be worth it.

Matt

Zephrant · #**107** 09/02/2003, 02:28 PM

NF- Same float switch is fine, and wire it in series with the other one. So go power, float, float, solenoid, neutral.

Basically cut a wire coming out of the float, and wire the new float switch in to each end.

Matt- I think that what you propose would work, but that doubles the current going though the little float switches. I don't think you add much more of a "safety factor" by adding the additional solenoid.

Zeph

RicoJ · #**108** 09/08/2003, 05:44 PM

Are there any picture of how this set up works with a kalk reactor?

Thanks!

Rico.

NoFear29 · #**109** 09/08/2003, 08:57 PM

Rico,

Try this link. I have some pictures of my setup towards the end of the thread.

http://archive.reefcentral.com/forum...5&pagenumber=1

Eric Wolfarth · #**110** 09/11/2003, 11:13 AM

Folks -

A word of caution, I actually spoke to the company regarding the float switch that you provided here. They said that the switch is epoxy sealed, but does not recommend that it be submersed, since the water will eventually make its way in. Not very promising. Although I have read a lot of threads, and no one speaks of this concern.

Anyways, be careful if your design i remains underwater.

A GFI is a must.

NoFear29 · #**111** 09/11/2003, 04:40 PM

A float switch that is not supposed to get wet. What a great idea. I do not remember anything mentioning that in the instructions.

NF

Zephrant · #**112** 09/11/2003, 10:34 PM

Thanks for the info Eric-

I guess the thing to do would be to silicon it in to a threaded bulkhead, with the wires outside the tank. Then they could not get wet.

Zeph

Eric Wolfarth · #**113** 09/12/2003, 09:36 AM

Zeph -

I inserted my switch into a 1/2 threaded "Tee" PVC, with some piping on each side of the tee. The tee/switch is positioned at the water top off line. The wires run thru the tee, and up thru the top section of the piping and out above any potential water line. The lower section of piping, I filled with sand, to insure that unit would remain upright in the water.

Zephrant · #**114** 09/12/2003, 01:58 PM

Sounds good- Is it affixed to the tank edge so a industrious turbo snail can't bull-doze it around?

Zeph

Eric Wolfarth · #**115** 09/12/2003, 02:35 PM

yes it is. however, this fills the loss of evaported water in my sump/ref.. I have no snails in my sump/ref, only macro, pods and bristleworms.

rallendorph · #**116** 09/14/2003, 10:28 PM

Zephrant - esteemed others...

For my Auto Top-off, I am currently using the Kent Float valve to control output from a Maxima RO/DI unit to my Fresh water holding bin (44Gal Rubbermaid container). I utilize a dosing pump (LiterMeter III) to add RO/DI water to my sump.

What I want to accomplish is a high-water (shut-off) to the RO/DI, and a low water (start) to the RO/DI. The low-water sensor would be at ~ 10 Gal level of the 44Gal Rubbermaid container, and the High-water just below the existing Kent Float valve. By doing this, the RO/DI unit will only be turned on when tank is near empty (10 Gal remains), and will produce ~30 Gal at one time. Which should provide me maximum benefit / efficiency from the RO/DI unit.

I may have read the answer to this in the thread, I'm not sure. Any thoughts how I'd put this together ? I'm having trouble figuing out how to design the two sensors together to allow me to only fill water when I hit the low-water mark.

BTW - My multiple Fail-safes includes an overflow bulkhead just above the Kent Float Valve. So a flood would require a stuck solenoid in RO/DI, the Kent float valve to be faulty, a second stuck float valve in my waste sump, and /or a faulty pump in the waste sump. n+1 redundancy levels. God I need a life.

Rob A.

Zephrant · #**117** 09/14/2003, 11:55 PM

The short answer is you need a relay.

Your power runs from the wall, though the top one (float-up=off), though the bottom one (float-down=off) and to your pump.

Now for the relay. It has the active terminals go across the bottom float, and the coil terminals from the wire between the top and bottom floats, with the other side going to neutral.

Here is how it works- The bottom float closes when you need water, when it closes the relay clacks closed too. Water fills the tub, and the bottom float switch opens which has: no effect, because the relay is bypassing that float now (latched it on). The tub keeps falling until the top float goes up (off) breaking the circuit, which allows the relay to open.

Now the water falls (evaporates), the top float closes, but nothing happens because the relay is open, and the bottom float is open too. When the water falls below the bottom float, it closes, the relay latches on, and the cycle continues.

It sounds complicated, but it is not too bad. Just get an AC relay and you are set.

Zeph

rallendorph · #**118** 09/15/2003, 12:31 PM

Zephrant -

Thanks. That's what I needed.

What's your current view of the solenoid your using ? If you were to replace it (or starting new), would you go back to the same product, or search for a different solution ?

Rob A.

Zephrant · #**119** 09/15/2003, 11:22 PM

Rob- I would spend a night or two searching for another solenoid. I'm not too worried about the one I use, but now that I know more about it, I would try to find something better.

Zeph

Team Slug · #**120** 09/16/2003, 12:33 PM

I made a top-off a couple months ago that works exactly as you are describing.

Make sure your relay is a DPDT. I used an AC coil because I didn't have room for another DC adapter or a separate transformer. In retrospect, I think I would have made room and gone with DC coil as it is very easy to provide arc suppression to the switch contacts (not so easy with AC coils). You can use an AC or DC coil vesion, it is up to you what you are more comfortable with.

Zephrant · #**121** 09/16/2003, 11:25 PM

Sounds good TS- Any chance you have a circuit diagram, or some photos you could toss up?

I'd like to see your ideas on arc suppression too.

Zeph

rallendorph · #**122** 09/16/2003, 11:27 PM

Thank you folks. I appreciate the advice.

Rob A.

Team Slug · #**123** 09/18/2003, 09:54 PM

Here's the schematic I used. It's pretty simple, but effective. It would need to be modified if you were to choose a DC coil on your relay instaed of AC.

I have been playing around with some new electronic switches from ST that are for switching AC loads (inductive or resistive) and require no snubbers or surge protection unlike traditional triacs. This in a nice small package that can handle up to 8A (althought that would be overkill for most of our applications). I have the 1A and 2A version and they seem to be working like a champ. Oh yeah...the best part is that they are switched on by a DC source. They can be driven directly by micro.

Zephrant · #**124** 09/18/2003, 10:03 PM

Nice schematic, thanks!

The switches sound good too.

Zeph

Gudwyn · #**125** 10/01/2003, 01:21 PM

I built one of these things with the relay. Used two float switches from mcmaster, one solenoid from mcmaster, 12v DC transformer, 12v DC reed style relay from radio shack, and a 12v LED from radio shack.

Solder jumper from one side of relay coil to the NO pin. Effectively there are now only 3 outputs from the relay. Then use wire ties to join together the rest of the parts.

To hold the float switches, I created U shaped loop from half inch PVC. Along the bottom of the U are two slip/threaded tee fittings. By angling one tee upwards and one tee downwards, I created about a 2" gap between the active points on the two float switches.

If I can convince my wife to take some pics I'll post em tonight