Help with a simple water level monitor please!

Thread Starter

mad__llama

Joined Jun 5, 2008
4
Hi All,
Please help out a complete beginner. I'm trying to build a very basic water level monitor but I think I'm trying to make it more complex than needs be.

To explain the "why": I'm a researcher at a University in Australia (Department of Biochemistry). Part of my job is to purify protein samples using a machine called an FPLC (basically a very expensive liquid pumping system). It draws liquid in from two sample bottles and pumps it through the system and various other components. Despite costing an insane amount of money (like all biomedical equipment) it does not have any way of monitoring when it has sucked up all the liquid in the sample bottle. This results in air being introduced into the system. This is not good. I'm pretty busy doing various other tasks around the lab and its very easy to lose focus and forget to continually monitor the water level.

So basically I'm trying to build a small widget which I place into the bottle that will sound an alarm when the liquid falls below a desired level so I can come running and replace the liquid or stop the pump.

My idea is based on a float type system: ie. a fixed rod with a small switch at the bottom which is inserted into a plastic test-tube and suspended in the water. While the water level is above the desired limit, the testube will float and press up against the switch at the bottom of the rod (or complete a circuit etc). When the water level falls, so will the test-tube, depressing the switch (or breaking the circuit). At this point I want an alarm to sound.

My thinking suggests that I need an inverter (NOT gate?) of some type so that while the switch is depressed, the alarm is silent but all the resources I've look at on the net seem to require a more-than-basic understanding of electronics.

So please help me out here (and contribute to biomedical research!). Any ideas which I might be able to implement would be great (the simpler the better). I'd also prefer to try and buy pre-made components rather than trying to put together complex circuits myself.

Thanks in advance for the help,
Mad__llama
 

mik3

Joined Feb 4, 2008
4,843
There are switches, called limit switches, which have three contacts. The contacts are called common (C), normally open (NO) and normally closed (NC). When the switch is not pressed a closed circuit between the C and NC pins exists and an open circuit between the C and NO pins exists too. When the switch is pressed a closed circuit exists between C and NO pins and an open circuit exists between C and NC pins. Thus you have to connect your circuit to C and NC and when the test tube wont press the switch it will fire your alarm or what ever.
 

jpanhalt

Joined Jan 18, 2008
11,087
When I was doing that type of work, I knew my reservoirs contained 1L and the pump rate was, say, 2mL per minute. I just set a timer. John
 

Wendy

Joined Mar 24, 2008
23,415
There are liquid sensors out there that use optical pickups, but are digital in nature. While complex on the inside (I'm sure) they are pretty simple electrically speaking. They use a cone that changes it's coefficient of reflection, when it's wet the LED light disperses into the liquid, when it's dry (or not immersed) the light bounces back to a sensor. Is this something you might be interested in? My experience was with machine maintenance, we used it to trip a timed pump to empty a reservoir. They are fairly largish, about 3/4 in diameter and and inch and a half long, as memory serves.
 

beenthere

Joined Apr 20, 2004
15,819
If your sample bottles are of small volume (I'm only familiar with HPLC), perhaps you could rig a microswitch that will trip when the weight on it gets low enough. A balance beam setup could be set up, so the one end goes down onto a switch and stops the experiment.
 

Thread Starter

mad__llama

Joined Jun 5, 2008
4
Thanks all for all the help so far!

When I was doing that type of work, I knew my reservoirs contained 1L and the pump rate was, say, 2mL per minute. I just set a timer. John
You'd think it would be that easy John :)

Unfortunately the indicated pump rate rarely coresponds to the actual pump rate. When I'm working with large volumes I have to take into account the 10-15% error in the machine. When I'm trying to suck up *all* the liquid in the reservoir without taking in any air, this means coming back every minute to watch the damn machine. Very annoying.
 

SgtWookie

Joined Jul 17, 2007
22,230
I really like Beenthere's suggestion.
It has the big advantage of avoiding introduction of possible contamination into the sample bottle, along with being quite simple.
 

Thread Starter

mad__llama

Joined Jun 5, 2008
4
Hi mad llama
Are the bottles clear glass? What size? Opening size?
The bottles are clear, usually 250 or 500 ml Schott bottles (standard lab size) with a 4 cm opening size. However I've already got the tubing from the pump going into the bottle so that restricts the available space.

I quite like BeenThere's idea except for one problem: Each bottle weighs has a slightly different mass and I use different volumes of liquid each time so I'd have to calibrate the sensor every time I use it.

I'd prefer to have a system I can just suspend in the bottle and quickly adjust to the liquid level I need.

This might seem silly but could I just suspend two wires in the liquid and use the liquid itself to complete the circuit? When the level drops below the wires, the circuit is broken and the alarm alerts me. This was my original idea but I rejected it for two reasons:

1. Most solutions are salty and I just assumed the wires would rust at some point (and possibly contaminate the solution).

2. The FPLC pump very accurately measures the salt concentration of the solutions it pumps (I guess by passing a current through it and measuring conductivity?) and I wouldn't want my circuit to somehow mess with that.
 

SgtWookie

Joined Jul 17, 2007
22,230
I suggest that any physical object entering the sample container could interfere with the quality of the evaluation.

However, if it's sterilized in the same manner that the tubing (etc.) is undergoing between samples, it might be do-able.

If the bottles are clear and cylindrical in shape, a low-powered laser (like a common pointer) could be directed off-center at one side of a sample bottle, with a photodetector on the other side. Light would be deflected when the water level was higher than the laser, but would pass nearly straight through and strike the sensor when the water level in the container was below that of the laser and sensor. However, this doesn't provide for a "fail safe".

In a "fail safe" mode, light would be refracted through the liquid in a diagonal to the target sensor. The target sensor would then monitor the light level; if there were either a decrease or increase in laser light detected, the alarm could be sounded, and/or the sample processing instrument/pump turned off.
 

Externet

Joined Nov 29, 2005
2,202
Hello.
If a layer of sterile, rinsed polyestirene spheres of about 3-4 mm diameter (those used as beanbags filler) is put to float on the fluid, an optical sensor can be placed on the jar exterior.
Detection of light (or pointer as Sgt says) from the other side of the bottle would be interrupted by the floating layer when level descends to the chosen location.

In other words, an opaque floating material is easily and reliably detectable, and the optical sensor can be any-time height-adjusted as desired from the exterior.

Miguel
 

thingmaker3

Joined May 16, 2005
5,083
If the electrode idea is deemed viable, I suggest electrodes from the Australian version of series 316L stainless steel. This alloy is highly corrosion resistant, and is used in many medical applications in the US. It is one of the so-called "surgical stainless steels."

I'll try to find an Australian cross reference alloy.
 

jpanhalt

Joined Jan 18, 2008
11,087
A Hastalloy is another option. Some protein work is very sensitive to metal ions. It would be important to know what the pumps are made from. I have seen systems in which the wetted parts were essentially Teflon and glass.

I am still puzzled why the OP needs to use every last drop of the mobile phase. Is the mobile phase degassed or open to the air?

John
 

Brass

Joined Oct 13, 2006
14
Another possible solution would be to have two sets of balance scales each with weight equivelent to the sample containers with the fluids on the opposite sides , beside each container place a magnet and use a hall effect switch to turn off the pump when the magnet comes in contact.

Thus the only calibration required would be weight changes with different sized samples and if you batch them together so that your doing all same weighted samples in a series there shouldn't be that many changes.

This would be a BTN GID solution and sometimes SIB.

Better Than Nothing
Get It Done
Simple Is Better.
:)
 

Thread Starter

mad__llama

Joined Jun 5, 2008
4
I am still puzzled why the OP needs to use every last drop of the mobile phase. Is the mobile phase degassed or open to the air?

John
I'm using the system to load a protein-containing liquid (usually 100-500ml) into the pump and onto an affinity resin column. This mixture is not degassed and is open to the air.

Once this protein containing liquid has been loaded (ie. every last drop has been sucked up by the pump), the tubing is moved to another bottle containing the mobile phase. This is why I need an alarm of some sort to let me know when all the mixture has been loaded (or is about to run dry).
 

jpanhalt

Joined Jan 18, 2008
11,087
You could use a bubble detector in the line from the protein reservoir that would stop the pump. Then with a Y-valve, connect the regular mobile phase to the pump. Air would never enter the pump. All sorts of things can be used to detect bubbles. One method is to use a vertical U-tube with a light beam and detector across the apex. If you have any old RI or other detector (UV, visible, etc.) that would allow sufficient flow rate, they could be in-line, and any bubble will cause a sharp change in the signal, which would activate your switch. Finally, if you are anywhere near a blood bank, dialysis center, etc. bubble detectors are used with some of the infusion equipment. They may give you an old unit.

John
 

beenthere

Joined Apr 20, 2004
15,819
We made pretty good bubble detectors with spring clothespins. Placed an IR LED on one leg and a phototransistor in the other (just stuffed into undersized holes). A bubble makes a big change in the phototransistor output.
 
Top