unlatching a latching circuit

[Hawkdl2]

I designed this latching circuit for my RO-DI system based on all the nicely done, but over my head electrical circuit diagrams I found on several boards. It seems to work on my bench. What I can't figure out is where to install a momentary switch in the system so I can restart RO-DI production before the water level gets to the lower float. Perhaps one of the electrical wizards out there can give me some tips.

 

[BeanAnimal]

Use a NO switch in parallel with the lower float
You can use a NC switch in series with the upper float as a kill switch.

(I can not check your schematic..., as I don't know what the pinout of your relay is but it looks right, as long as the center horizontal lugs are COMMON and the uppers are NO)


You really should be using a diode to protect the float switches from the inductive kickback produced when the relay coil is de-energized.

 

[driftin]

Bean, where would the diode go?

 

[BeanAnimal]

 

[Hawkdl2]

Bean,

Thanks. I'm sure I'm not alone, but I have a heck of a time translating schematics into a "what connects to what" diagram. My relay is the one from Autotopoff. Any chance you could suggest where the diode goes based on my diagram?

 

[jdieck]

The diode goes across the terminals of the relay's coil to prevent transients from having the relay "hesitate" It must point from the negative to the positive. In your diagram it will go conecting the two top terminals in the drawing (the coil) pointing towards the left terminal (The one conected to the float).
You can also conect a 12 volt LED across the same terminals to know when the relay is activated (I use red).
I also installed a green LED across the + and - of the 12 V power supply to know when there is power from the adapter.

To activate it manually before the level reaches FS1 (Bottom level switch) install a normally open (NO) momentary push button in paralell conected to the terminals of FS1 (lower float). Pushing the button will activate the relay and memory contact but only when the level is below FS2. Once the level reaches FS2 it will deactivate.

 

[jdieck]

^ Oops!

 

[BeanAnimal]

Please don't take this the wrong way... but a schematic is a what connects to what diagram. A schematic is a universal language used to communicate the structure of electrical circuits.

You are DIYing with electricity, so it is rather important that you take the time to understand exactly what you are DIYing At the very least, learning to read basic schematics will help a lot.

In this case you should be able to easily identify the parts in the two pictures and label the wires. A process of elimination would be a good start. I have done it for you with color.

Notice the RELAY in the schematic has (8) connection points, just like the one in your picture. It has (2) connection points on each float switch. With a bit if logical we can map out all of the connections and see that they are exactly the same. We can ignore the 120V side, as you are not modifying it at all.

 

[BeanAnimal]

<a href=showthread.php?s=&postid=15244840#post15244840 target=_blank>Originally posted</a> by jdieck
The diode goes across the terminals of the relay's coil to prevent transients from having the relay "hesitate"

Actually, the diode will slow the relay coil down (cause it to 'hesitate') slightly

When the coil is de-energized the magentic field holding the armature collapses the spring sucks the metal armature back through the coil. This motion (and the field collapse) causes an electric current to be produced (the same as a generator). This induced current can damage upstream components. The diode feeds this current back through the relay coil and dissipates it.

We use different protection schemes depeding on the application. In a nutshell, the higher the level of protetion, the slower the action of the relay and the larger the arcing at the contacts. The trick is finding a happy medium

For our purposes, it does not matter much

 

[jdieck]

Actually I had a diode installed and removed it given that in the circuit there is no sensitive electronics.
Have operated fine for a couple of years now, I think some of the capacitors in the power supply may be dampening the transient.

 

[Hawkdl2]

Bean, no offense taken at all and I appreciate the Rosetta Stone. your comments are very helpful and well taken. Also, I grossly overstated my ignorance (I think). I created the easy-to-follow diagram based on yours and others' proper schematics. When I'm in the shop, however, I find my reductionist diagram makes remembering to cut the blue wire instead of the red wire a lot easier.


So, do I understand it correctly, the consensus is no diode?



jdieck, thanks very much - that was helpful.

 

[BeanAnimal]

<a href=showthread.php?s=&postid=15245331#post15245331 target=_blank>Originally posted</a> by jdieck
Actually I had a diode installed and removed it given that in the circuit there is no sensitive electronics.
Have operated fine for a couple of years now, I think some of the capacitors in the power supply may be dampening the transient.

Yeah the transients can be hard on ICs and other sensitive components. The reason I like to see then on these circuits is due to the reed switches in the floats

 

[BeanAnimal]

Hawk... You will likely be fine without the diode, but they are best practice and will (in theory) decrease the chances of your circuit failing

 

[jejton]

I have no comment on the electrical as it is all beyond me but I noticed that you have it set up in the GIF that FS1 turns the ATO pump on when it is lowered but does not turn the pump off when it is raised, and rather only FS2 turns the pump off when it is raised. Wouldn't it be safer to have FS1 also turn off the pump when the water reaches it and only have FS2 as a backup in case FS1 fails to turn off the pump?

 

[jdieck]

<a href=showthread.php?s=&postid=15246494#post15246494 target=_blank>Originally posted</a> by jejton
I have no comment on the electrical as it is all beyond me but I noticed that you have it set up in the GIF that FS1 turns the ATO pump on when it is lowered but does not turn the pump off when it is raised, and rather only FS2 turns the pump off when it is raised. Wouldn't it be safer to have FS1 also turn off the pump when the water reaches it and only have FS2 as a backup in case FS1 fails to turn off the pump?

Hawkdl2:
Install the diode as a best practice. The quality and capacity of the reed switches vary so it is better to have it than not. Mine is just a practice run but I use industrial strenght floats.

jejton:
The shown circuit turning on and off the RO/DI supply is basically to be used in a reservoir tank where FS1 is installed a tthe bottom (Turn ON) and FS2 is installed at the top (Off) so there is a relative large batch of water produced. I would not use it for a sump top off as the amount of water per batch will be too small and no purity of the RO/DI will be achieved with every start up.
As a safty, I would install a float valve at the point were the RO/DI enters the reservoir and installed at a point higher than FS2 so in case FS2 fails to stop the RO/DI, the float valve will close , the pressure in the product water line increase and the ASOV valve in the RO/DI unit will stop the feed (and waste) of the RO/DI unit.

If I would use this circuit for sump top off. I would place both floats close to each other but still FS2 higher than FS1 and I would have it run a powerhead or aqualifter from the reservoir (Not RO/DI) into the sump and again my recomendation will be a float valve above the level of SF2.
Having a single float turn on and off the pump will call for an unestable system as any small waving or changes in the level can have the float turing on and off almost constantly rattling the pump with no delay time between on and off.
Having said that, there are floats designed to do that that have a longer inner shaft and it takes some level change beteen the on and off positions.

 

[BeanAnimal]

<a href=showthread.php?s=&postid=15246494#post15246494 target=_blank>Originally posted</a> by jejton
I have no comment on the electrical as it is all beyond me but I noticed that you have it set up in the GIF that FS1 turns the ATO pump on when it is lowered but does not turn the pump off when it is raised, and rather only FS2 turns the pump off when it is raised. Wouldn't it be safer to have FS1 also turn off the pump when the water reaches it and only have FS2 as a backup in case FS1 fails to turn off the pump?

This is a classic latching circuit. It is designed to fill a resevoir by cycling a pump (or other device) on and off, not maintaining a set level.

As jdieck mentioned, a second "fail-safe" float should be added IN SERIES to the power supply and set ABOVE the upper float. This will ensure that a failed float or welded shut relay contact does not overfill the system.

Actually, I prefer to use a mechanical float valve as the actual endpoint of the water feed. Set above the top float (and/or fail-safe float) it will mechanically shut off the water in the event of a failed open solenoid (or runaway siphon).

 

[jdieck]

Here you can have a better idea of the location of float switches and float valve.
The yellow line is the withrawal to the doser pump feedign the Kalk reactor to replace evaporation.
Although not aparent in the picture the float valve is at a higher point than the higher level float switch.

 

[BeanAnimal]

How utterly scary.... but it looks like you secreted into my basement and stole my mixing container!

 

[jdieck]

Scary? May be just forgetful
Hey Bean, remember a couple of years back you asked for how my system was set?
Actually that is an older picture, now I have replaced the Brute with a graduated 55 gal poly tank as I needed a lot of water when I replaced the main tank.
I think we even had a discussion about the diod, something about if it was better to use a diode, or adding something else (resitor, capacitor, inductance) heck I do not rememvber well but something on that line.

Remember this drawing?

 

[BeanAnimal]

Don't remember the drawing, but do remeber talking to you about your setup or relays or something. It was not long ago when only a few of us fish geeks knew what a latching circuit was. Kind of funny that several ATO companies have sprung up to sell the exact circuits that we draw up for people (not that there is anything special about a latch). The best part is that one of them copied an untested concept type schematic that I posted here on RC. They did so without asking or understanding the circuit topology and its serious flaws. They sold the unit for at least a year with the severly flawed "run dry" protection.

Anyway...