# Homework Help: Non-linear conductivity of tap water

1. Jul 12, 2016

### Roger44

Hi everbody
My plots seem to show that conductivity between two nails immersed a few mm in tap water falls off at low voltages instead of being a horizontal flat line as would be the case for a linear ohmic medium. And the deeper the immersion, the less linear the conductivity. Is this normal or an experimental error?

I measured 50Hz AC current between two stainless steel nails in cold tap water in a small round dish at 5 different voltages, and repeated at five depths (a few mm)

2. Jul 12, 2016

Interesting data. I do think at lower voltages you might see something like a "contact potential" phenomenon where you get nearly zero current unless you exceed a certain voltage. There could also be a voltage amplitude dependent phase term caused by the time it takes for the charges (electrons) to travel from one nail to the other.

3. Jul 13, 2016

### Staff: Mentor

Water is not a linear ohmic medium.

4. Jul 16, 2016

### Roger44

Hi

There was an elementary shame-on-you experimental error. My 50 Hz AC supply was a simple "about" 10V output mains transformer whose voltage fell off when I pulled current. Temperature variation imprecision too. I dug out my signal generator, checked it on the oscillo' and started again. There's experimental imprecision at low values.

5. Jul 16, 2016

It's interesting how "water" which would not generally be categorized as something that would necessarily obey Ohm's law seems to follow it fairly closely.

6. Jul 16, 2016

### Roger44

Now the real challenge would be to deduce by scientific reasoning why simple tap water is linear at these voltages ranges and wouldn't be at other temperatures/concentrations/voltages, if that is the case, Beyond me, friends!

7. Oct 8, 2016

### Staff: Mentor

So you're now distancing yourself from your earlier results where resistance appeared to vary with depth over just a few mm.?

8. Oct 9, 2016

### Roger44

Hi
No, resistance is a very good indicator of shallow water depth variations but I'm not sure you can calculate exactly the depth because the voltage/current may not be linear. I'm still working on this possible non-linearity.

Firstly I'm awaiting a precise AC signal source which should counter the erratic values at low voltages, at the bottom left end of the curve. A second problem is the fact that when current flows it heats water and hot water, unlike copper, conducts more. So there's a sort of thermal runaway until thermal equilibrium.is reached. This would account for the rising slope.
I've measured under flowing constant temp tap water with a thermistor between the electrodes. The thermistor still reports a temp increase, but less than a °C. The slope is still rising to an extent that cannot be accounted for by such a small temp increase.

I'll try to push the experimental precision further by measuring in the stream at the bottom of the garden in order to have an abundant gush of constant temp water, and a constantin thermocouple to have temp changes according to a different technology.

I've also recorded the very first value displayed by the voltmeter when the circuit is closed, before the water begins heating. This also shows a rising slope very similar to the 'almost constant temp' slope. So conductivity may not be linear. I shall improve the rapidity of the measure by a 10 readings per sec digital recording.

All comments and suggestions are more than welcome. What bothers me is that nowhere on the Net can I find any mention of non-linearity of electrolytes. Even manufacturers of conductivity cells for measuring properties of liquids make no mention of it. Lots and lots on temperature non-linearity.