Measure water conductivity/resistance

Hi :slight_smile:

I need your help for a new project I'm starting:

  • I will put deionized water (18 Mohm water) in a tube. One electrode at each end. Is there a way to measure the tension between the 2 electrodes ?

  • then I'll add some salt. Let's say NaCl for now. But I'll add very few of the salt. The final concentration should be around 5 µM for example. Is there a way to measure the new tension ? Of course, the difference between before and after the salt should be measurable reliably.

Then I have a few question. I have seen plenty of conductivity probes. If I remember my classes correctly, what they do is to measure a tension difference, and then convert it to a Siemens/cm value, right ? So, what is the difference between that, and a voltmeter ? The sensitivity ?

For my project, I don't need to measure the real conductivity, I just need a value before/after, so a tension or a resistance would be ok. Also, the measure would be very short: measuring, and then switching the sensor off, so normally even oxydizable metals would fit my needs.
So, is there a way to get one of this value with a good sensitivity ? And cheaply, of course...:slight_smile:

Thanks !

Are these electrodes different metals? If they are the same, there is no voltage between them. Is that what you meant by "tension"?

For measuring conductivity, you use the same metal in both electrodes and apply an outside source of voltage and current. You may apply a constant voltage and measure the current, or apply a constant current and measure the voltage drop. This is basically just an ohmmeter.

You could, in fact, simply connect a DMM set on Ohms, although you'd have to make sure the resistance scale goes up high enough for what you are measuring. Most cheap meters only go to 10M or 20M ohms maximum.

You do know that Siemens of conductivity is just the reciprocal of Ohms of resistance?

To get it as Ohms or Siemens per cm, you’ll need to control the size and distance of the electrodes.

And you’ll need to know the temperature of the water, as that affects conductivity.

http://www.mbhes.com/conductivity_measurement.htm

polymorph:
You could, in fact, simply connect a DMM set on Ohms, although you'd have to make sure the resistance scale goes up high enough for what you are measuring. Most cheap meters only go to 10M or 20M ohms maximum.

Yeah you'll maybe need a kick-ass Megger

Actually, I'm a chemist, so the conductivity part, I know it. What I do not know is the "electronic part".

@polymorph:

You do know that Siemens of conductivity is just the reciprocal of Ohms of resistance?

To get it as Ohms or Siemens per cm, you'll need to control the size and distance of the electrodes.

And you'll need to know the temperature of the water, as that affects conductivity.

That's what I need to do. But I don't care about the value in siemens. Let's assume the size and the distance of the electrodes won't vary (inert electrodes, no erosion, or calibration, whatever). And let's assume (for now) the temperature is always the same.

For measuring conductivity, you use the same metal in both electrodes and apply an outside source of voltage and current. You may apply a constant voltage and measure the current, or apply a constant current and measure the voltage drop. This is basically just an ohmmeter.

Yes, that's what I need to do. Which way is the more sensitive ?

You could, in fact, simply connect a DMM set on Ohms, although you'd have to make sure the resistance scale goes up high enough for what you are measuring. Most cheap meters only go to 10M or 20M ohms maximum.

:slight_smile: I'm sorry, what is a DMM ? And if I understand well, the higher the resistance will be, the higher the sensitivity ?

DMM = Digital Multi Meter like one of these.

I think the sensitivity will depend on the make and model of the meter.

Ok.

Yeah, as the starting resistance would be 18 M ohms, 20 M ohms is mandatory. But can’t I do that with an Arduino ?

But can't I do that with an Arduino ?

Yes, but you will need external circuitry to measure the very low currents expected.

Ok. What would I need ? Do you have any lead about how to do that ?

I'm pretty sure that Mr. Google could help with that question.

Did you read the link I gave you?

Probe Types and Polarization Errors

Amperometric

The probe used to measure conductivity was originally an amperometric system which had two electrodes spaced one centimeter* apart from each other. [* Probes with different electrode spacing allow measurement of various conductivities.]

The amperometric method applies a known potential (voltage, V) to a pair of electrodes and measures the current (I). According to Ohm’s law: I=V/R where R is the resistance. The higher the current so obtained, the greater the conductivity. The resistance in this method unfortunately is not constant even though the distance may be fixed. Salt deposition on the electrodes due to electrolysis can vary the resistance. For low to medium levels of conductivity (< 2 mS/cm) this may be sufficient, but for greater accuracy and for higher levels, a different method is required.

Potentiometric

A potentiometric method is based on induction and eliminates the effects of polarization common to the amperometric method. The potentiometric method employs four rings: the outer two rings apply an alternating voltage and induce a current loop in the solution while the inner rings measure the voltage drop induced by the current loop. This measurement is directly dependent upon the conductivity of the solution. A shield around the rings maintains a constant field by fixing the volume of solution around the rings.

Because a potentiometric (4-ring) conductivity sensor is not limited by electrolysis which commonly affects amperometric probes, it can be used to measure a much wider range of conductivities. Practically, stainless steel rings can be used. But, the preferred metal is platinum because it can withstand higher temperatures and produces a more stable reading. Platinum sensors are also easier to clean. Advanced microprocessor conductivity instruments can vary the voltage applied to the sensor which enables them to extend the range of a potentiometric probe even further. This technique allows advanced meters to be able to measure both high and low conductivities as well as the ultra low conductivity of deionized water with one probe.

Inductive or Toroidal

Another method of conductivity measurement uses an inductive probe (sometimes referred to as a toroidal sensor). Typically these are found in industrial process control systems. The sensor looks like a donut (toroid) on a stick. The advantage of this technology is measurement without any electrical contact between the electrode and the process fluid. The probe uses two toroidal transformers which are inductively coupled side by side and encased in a plastic sheath. The controller supplies a high frequency reference voltage to the first toroid or drive coil which generates a strong magnetic field. As the liquid containing conductive ions passes thru the hole of the sensor, it acts as a one turn secondary winding. The passage of this fluid then induces a current proportional to the voltage induced by the magnetic field. The conductance of the one turn winding is measured according to Ohm’s law. The conductance is proportional to the specific conductivity of the fluid and a constant factor determined by the geometry and installation of the sensor. The second toroid or receiving coil also is affected by the passage of the fluid in a similar fashion. The liquid passing thru the second toroid also acts as a liquid turn or primary winding in the second toroidal transformer. The current generated by the fluid creates a magnetic field in the second toroid. The induced current from the receiving coil is measured as an output to the instrument. The controller converts the signal from the sensor to specific conductivity of the process liquid. As long as the sensor has a clearance of at least 3 cm the proximity of pipe or container walls will have a negligible effect on the induced current.

As you can see, they go in order of complexity from simplest to more complex, and from least reliable to most reliable.

However, the toroidal method works better with high conductivity solutions. It sounds like the potentiometric would be better for your purposes.

http://www2.emersonprocess.com/siteadmincenter/PM%20Rosemount%20Analytical%20Documents/Liq_ADS_43-018.pdf

It is basically an AC 4 point measurement. The equivalent in measuring resistance in a wire would be Kelvin probes measuring low resistance, a way to take the contact resistance out of the equation.

Now… what part do you need help with?

Hm. That last line isn't meant to be as aggressive as it may appear.

Hi polymorph,

very interesting the documentation in the link you posted. But the problem is how to measure de resistance in AC and interfacing to the arduino? for what you uses the demodulator in this case? because we are not measuring inductance.

here there is a model for the measurement wiht two electrodes but for me the problem is the same: measure resistante or voltaje in ac and interfacing to the arduino when these are too low or its range is high variability.

any ideas for the “read out device” in the range from 1khz to 20khz.

anyway thankyou for your answers!! :smiley: