Reservoir Waterlevel Schematic

[redleader]

How about an IR detector and an IR sensitive dye? Not sure how easy it would be to get the dye right, but IR LEDs are pretty common.

[JSimmons]

Hence my original post concerning opaque water.

I have a bay-res reservoir, but it's dimensions don't fit the intended location very well, and there's the issue of the "filler cap possibly leaking (although that's easy to fix).

So, I was off looking for a different kind of reservoir. At first, I thought I'd try one of the 6-inch criticool waterplant reservoir, but my local electronics expert and I agree that the arc of the round reservoir would probably have some bizarre and unexpected effect on the IR light beam.

Right now, I'm considering making a custom reservoir that looks something like the drawing attached to this message.

The idea is that if I make a separate sensor trough, the water won't be as agitated as the water in the main reservoir bay, so the sensor readings will be more stable.

Also, I can probably reduce the intensity of the beam enough that it won't penetrate water that is mixed with some sort of semi-opaque substance.

[the creature]

Originally posted by zoson
Electricity through water = short to ground through radiator (depending).
electricity through water = copper waterblock conducting electricity on top of your cpu core.
-Zoson

Agree, If I had built somthink like that I would use capasitans to mesure the levels, but diffrent setup diffrent water-mass

[thoro]

how about using IR ranger dectectors? coming from above, bouncing off of the floater. see image, i can't describe.

[JSimmons]

We've alread considered that, and aren't sure if we can get anything that measures such small distances.

We always keep coming back to the IR emitter/sensor stuff.

[thoro]

ok, maybe i just missed that reply. with finals next week............ brain is not working

[bigben2k]

I don't see a problem here.

You're not limited to IR, you can use any visible light to sense the level of the float.

[JSimmons]

Originally posted by thoro
ok, maybe i just missed that reply. with finals next week............ brain is not working

By "we", I meant me and the guys at work who are co-pondering possible solutions.

[TerraMex]

>Also, I can probably reduce the intensity of the beam enough
>that it won't penetrate water that is mixed with some sort of
>semi-opaque substance.

@Jsimmons

You dont need that many emitters/receivers. 4 or 5 will do. But you need to plant them in the opposite wall, at the same level. When the level drops, the IR beam gets across, and triggers the receiver. That with a opaq liquid.

http://clientes.netvisao.pt/carlo001/procooling/13.jpg , like this.

You can use different approaches to the same type of measure with light. You can use a powerfull light on top of the res and use an LDR cell on the bottom of the res. As the level drops, more light gets to the bottom and the resistance change can trigger anything you want. But this works better with a clear, or semiclear liquid (and a low ambient light)

Again, it's not easy to use a vertical approach to a IR beam. It bounces off the water and return to the receiver with very little drop in intensity. You need to measure the time and not the intensity to use that setting. A foamy, turbulent surface would invalidade any of those uses. Besides, you'd need to have a sensibility of (let's say) 0.1 inches in variations to get a decent measure.

Here's a couple of more ideias :

http://clientes.netvisao.pt/carlo001/procooling/08.jpg

Capacitance measurement. The rods are isolated so there is not actual contact with the water. The system measures the alteration of the dielectric coeficient of the water as level changes and produces an alteration to the capacitance. Connected to a Wheastone capacitance bridge , it can produce (with a filter) a direct relationship between level and signal amplitude.

http://clientes.netvisao.pt/carlo001/procooling/09.jpg

Pressure switches, similar to a relay. Something like this : http://www.grainger.com/Grainger/pro...mId=1611760105

http://clientes.netvisao.pt/carlo001/procooling/11.jpg

DC-DC LVDT movement measure. Operates by amplitude changes by the movement of the coil through the lenght of the coils. Can be used in my early description.

[bigben2k]

I updated my drawing: maybe it makes more sense now

[TerraMex]

@BigBen2k .

Right. That will work (if properly done). It doesnt have to be IR emittor/receiver pair . You can use a good range of light sources (like a standard led) and a phototransistor or an LDR resistor. A good set of leds in a plexi reservoir would have that "extra pimp" look to the case.

However it must be taken in account exterior light , and EMI to the circuit.

Btw, the buoy might not be necessary in an opaq coolant. The refraction of light might be sufficient to measure two independent levels of voltage amplitude that can be translated as in "in water" and "not in water" .

[the creature]

or.... why don't masuere the weight of the resovair, when it's emty it dosen't weight so much

or...... why don't use ultrasound, the sound is reflectet back to a sensor when it hit the water

or.... why don't go the easy route?

[TerraMex]

... Because you need a pressure sensitive sensor, and most are expensive or need a complicated assembley . You could get a piezoelectric one, but thats not easy to find .. over here anyway.

... Because it envolves a more complicated schematic, setup, and setpoints. And Ultrasonic is annoying with the clicks .

... easy route = exterior clear reservoir .

[JSimmons]

Well, I'm gonna use an interior pre-made reservoir until I can come up with a viable (and admittedly mostly useless) water level sensor system buil;t around a custom-made reservoir.

"Sure it's useless honey, but look at the pretty lights..."

[TerraMex]

Alot of what was discussed is doable. Use the light+sensor+bouy sensor type thing. Pretty lights, a resonably accurate level sensing and you can adapt to the initial schematic. Using a phototransistor you can control directly the output. It'd be like using it as a relay.

Or, use the LDR on the bottom of the reservoir and a few leds on top, and you'd have a variable resistance (use a wheastone brige for compensating) . As the level fluctuates , so does the light reaching the bottom. But that depends alot on the transparency of the fluid and exterior light (which if it stays inside the case, shouldnt be a problem).

Both of these look to me like the real "easy" ones.

[bigben2k]

a simpler method would be a set of electro-mechanical float switches, but we'd be back to the electricity in the water problem.

Why don't you attachd a magnet to the float, and use sensors to detect it? Nah...

The light detection is simple and elegant. Go for it.

[wymjym]

Here is a thought,
Rather than doing a really trick sensor (with all the associated fail safe, fault tolerant, etc circuitry) let us try a rather basic…electro-mechanical sensor.
Within the res add a small vertical chamber that is open at the bottom and top
Within this chamber place a small buoyant float assembly that has a magnet attached to it. (Could be encased in plastic if the 'fear' of metal in the water upsets you)
On the outside of the res attach small reed switches at various levels
Then use the reed switches to trigger whatever type of level monitoring you wish to use.
wj

[TerraMex]

As long as the magnet is isolated from the water.

Usually having a relatively strong magnetic field in water is a big "no no". It will eventually interact with the ions in the water and create a nice layer of magnetized compound around the magnet, twisting its fields . The way to prevent this is a electromagnet with an AC signal. And we'd be back to the electric current inside water.

But it might work. Using distilled water, and with a small magnet.And if not a Reed Switch, a Hall sensor could also work.

[wymjym]

Originally posted by TerraMex
As long as the magnet is isolated from the water.

Usually having a relatively strong magnetic field in water is a big "no no". It will eventually interact with the ions in the water and create a nice layer of magnetized compound around the magnet, twisting its fields . The way to prevent this is a electromagnet with an AC signal. And we'd be back to the electric current inside water.

But it might work.

as I said,
'Could be encased in plastic if the 'fear' of metal in the water upsets you'
mix up a little epoxy and drop the sucker into it, protected!!!
I'm not sure what a relatively strong magnetic field is, I was thinking along the lines of using a very small (weak) magnet since the reed switches I have require a minimum of pull to close them.

I assume that when someone says, it might work..within these forums...that should be read as 'it is more than likely that it will work'.




wj

[TerraMex]

You misunderstood (or i didnt get my point across). Anyway, the problem is not the metal in water, it's the magnetic field.

The initial testing with these types of sensors, magnetic displacement in flow measurement (to be more precise, but it can be interpolated to here), using a magnetic sensor, resulted in precipitation of ions on the surface of the magnet (which is isolated, but doesnt matter as long as the field exists), distorting the field, and giving inaccurate readings, or even reducing the field considerably , with some extra EMI , making the sensor useless.

However, as i said, distilled water with a low ions count, a magnet with a light field, and those reed switches , in theory, they WILL work. The whole issue is the remaining ions in the water, and the long term effects on the magnet . Using that setup and having the whole thing stop working properly in two weeks isnt what we want .

But "gopher it" , and give us some feedback . I'm having trouble locating some reed switches here for some testing. Damnit.

[wymjym]

I misunderstood, but
If I'm reading this correctly you are still suggesting a much more sophisticated and sensitive mechanism than I am. I envisioned a 5 inch tall container with a reed every 3/4 inches, with an 'active' reed field of about 1/4 inch. Use whatever circuitry is necessary (out side of the container) to give the visual effects.
Sorry but at present my dummying time is already allocated to other things.
wj

wj

[TerraMex]

No no no, not more complex. Same thing.
Just that you need to be carefull with the magnetic fields in water.

http://international2.farnell.com/Se...64&SKUID=36503
Like these?

[bigben2k]

How about mounting the magnet on a pulley, so that it can sit dry?

[JSimmons]

At this point, one of those little X10 cameras connected to a CCTV monitor would be okay.

Of course, it would have to be mounted on a little tram system so it could move around in the case.

[Alchemy]

Originally posted by TerraMex
The initial testing with these types of sensors, magnetic displacement in flow measurement (to be more precise, but it can be interpolated to here), using a magnetic sensor, resulted in precipitation of ions on the surface of the magnet (which is isolated, but doesnt matter as long as the field exists), distorting the field, and giving inaccurate readings, or even reducing the field considerably , with some extra EMI , making the sensor useless.

I'm not sure what qualifies as "initial testing," but I can tell you that I've done the same thing with ferromagnetic stirrer-bars in solutions of water and both cationic and anionic chromophores, and I've never gotten "ion precipitation" other than the expected EM bonding on the magnet's plastic casing (aka, the stuff stains.) How strong a magnet are you using, for god's sake?

Seriously, the sort of small permanent magnets you're going to use for this (~0.01 T) will *not* affect water that has a conductivity on the order of tap water. The only troubles I've seen with magnets this size are with fluids that are by themselves ferromagnetic - certain types of fluids for photographic film deposition, for example. I find it unlikely you are detecting evidence of the Hall effect unless you're strapping an NMR to your reservoir or have some very classy detection equipment. I find it *very* unlikely such an effect in this magnet could ever cause the magnetic field a centimeter from the magnet to reduce ten times over or reverse entirely, which is the only way such a simple solution as wymjym proposes could not work.

You're suggesting that the magnet wymjym suggests has such a *huge* charge flow and density about it that it can function as a galvanic cell? Where would the energy come from? You're also suggesting the layering that would be done on it would *distort* the magnetic field. How and why would a magnetic field arrange ions about itself in a pattern *contrary* to its magnetic field?

Look, I'm not an expert on magnets or electricity, but this whole discussion about how scary magnets are reeks like those new-age magnetic bracelets used to cure stupid people of having wallets.

http://education.magnet.fsu.edu/faq/index.html

Alchemy