Conductivty and temperature

[Hawkdl2]

I've been setting up and evaluating my Neptune Apex conductivity probe in support of my automated NSW system. It took some time to set up and calibrate the conductivity probe and I had to use a temperature compensation offset of +2.7 for to get the probe to read 35 ppt using both conductivity and salinity calibration solutions. In any case as you can see below, conductivity remained fairly stable for several days despite notable temperature fluctuation in my unheated NSW tank - located in my garage. However, after a few days, conductivity began fluctuating fairly significantly with temperature changes. I am a little surprised by this - expecting the temperature compensation to keep the reading stable.

Does anyone else use a conductivity probe in a non-heated holding tank?

http://i12.photobucket.com/albums/a213/hawkdl2/conductivityswings.jpg

[mscarpena]

Temp and conductivity should be independent from each other. You should read about water conductivity on the web, but it is basically the amount of "stuff" in water. The only way it would be true is if there were undissolved particles in your tank and when the temp raised up it dissolved then temp drops and it drops back out of solution.

[TheH]

This looks pretty interesting. I think conductivity probes are sensitive to temperature, but they are almost universally automatically temperature compensating.

Are you still observing this kind of fluctuation with temperature?

[BeanAnimal]

Temp and conductivity should be independent from each other. You should read about water conductivity on the web, but it is basically the amount of "stuff" in water. The only way it would be true is if there were undissolved particles in your tank and when the temp raised up it dissolved then temp drops and it drops back out of solution.

Conductivity is extremely dependent on temperature. As the temperature of a substance rises, the electrical conductivity decreases (unless the substance is a semiconductor, but that is out of the scope here). In many subtances, the relationship between temperature and conductivity is somewhat linear, in others it is not so linear. This is very basic physics :)

[BeanAnimal]

This looks pretty interesting. I think conductivity probes are sensitive to temperature, but they are almost universally automatically temperature compensating.

Are you still observing this kind of fluctuation with temperature?

There is no such thing as a temperature compensating probe :)

A conductivity probe is comprised of two electrodes at different potential. A current is passed from one electrode, through the substance to be measured to the second electrode. The resistance to the current is measured. This current is very tiny and needs to be amplified so that our "meter" or "controller" can read it.

The amplifier and/or software are the only things that can compensate for temperature fluctuations. In some cases, the compensation is very crude or non existent, in others it is very sophisticated.

[TheH]

I hadn't thought of that BeanAnimal. But still, the probe must have a temperature sensor on it so that would make it part of the compensation mechanism :)

[BeanAnimal]

I hadn't thought of that BeanAnimal. But still, the probe must have a temperature sensor on it so that would make it part of the compensation mechanism :)

Some do have temperature sensors built in, some don't. It is the controller or software that uses the information from the temp sensor (if available) to scale the conductivity reading. I just ordered TDS probes for a project, (2) with temperature sensors (2) without :).

[Hawkdl2]

I'm not completely certain if the probe Apex uses is self temperature compensation or if it relies on the optional temperature probe connected to the same module (I would guess that's the approach they use), but I do have the temp probe and I know the system allows for adjusting of a temperature compensation offset value. Higher end probes have two sets of electrodes with one used as in internal reference to compensate for dirty or otherwise compromised test article electrodes. I have no idea if the Neptune "lab" probe is a 2 or 4 electrode unit.

My lab uses conductivity extensively, but over narrow temperature ranges (room temp.). What I have no experience with is measuring conductivity over broad(er) and fluctuating temperatures - i.e. my garage.

Calibration for the probe has been challenging and certainly a lot harder than my lab instruments, but then again, those cost several thousand $$.

My question remains; does anyone else measure conductivity in an un-heated tank and see steady or fluctuating conductivity over temperature?