[mikgol]

Hi there,

I am building a dog water bowl that will detect water level and automatically fill up.

To detect water level I am passing a small current through the water and measuring resistance.

As this is for drinking water, I want to be certain that I don't poison my dog.  My first thought was to use stainless steel, but I later found out that it shouldn't be used as electrodes because it has a toxic chemical in it (Chromium).

My question is - what's the best "food grade" metal that I can safely use as my electrodes to pass current through drinking water? Should I use copper or aluminium? Or is stainless steel ok as long as the current that passes through is very small?

[Borek]

Stainless steel should be OK as long as there is no electrolysis taking place. For that keep voltage as low as possible (0.5 V will do) and - if possible - use alternate current for the measurements. Plus, you don't have to keep it on all the time, measuring resistance once each 10 seconds (or half a minute) should do.

And yes, low current is a good idea, but currents in electrochemistry are not exactly ohmic.

[mikgol]

Thanks for the reply Borek.  I'll use a voltage divider circuit to bring the voltage down as low as I can and try it out (I'm using a 3.3v arduino and reading voltage on the analog inputs).  I'll also just send a short "pulse" every 10 seconds (for a few ms).

I'll also design it so that the "pulses" alternate direction each time as per your suggestion.

From your post I'm understanding that electrolysis amount is more influenced by voltage rather than current - is that accurate?

[Borek]

From your post I'm understanding that electrolysis amount is more influenced by voltage rather than current - is that accurate?

Note exactly.

"Amount" is given by current (more precisely: by charge, that's what https://en.wikipedia.org/wiki/Faraday%27s_laws_of_electrolysis say). But the reaction won't take place until the voltage gets higher than some threshold value.

You may want to google for "conductivity measurements", there are some well established methods, no need to reinvent the wheel

[mikgol]

Thanks for the awesome info Borek, that helped heaps cheers mate.

I've come across graphite electrodes (http://www.sglgroup.com/cms/international/products/product-groups/gce/graphite-electrodes/index.html?__locale=en) - they look like they would be ideal in this situation.  I'll do as you suggest with the low voltage & current + graphite rods, then it should be sweet 

[Borek]

I am not convinced graphite electrodes will be sturdy enough - they tend to break apart in time.

[Intanjir]

You don't actually want to conduct electrons as that would be electrolysis. So you don't need a conductor touching water.
What you are constructing is a capacitor with water as the dielectric.
This works well as even pure water is an excellent dielectric as far as liquids go with a much higher permittivity than air.
Of course a capacitor needs a conductor to form the plates but you can coat those plates with insulator so no metal intentionally touches the water.
A low voltage A/C will pass with zero electrolysis.
So the choice of conductor can be replaced by a choice of insulator.

If you aren't electrolyzing than it is much easier to use graphite electrodes with water. My understanding is that one of the main problems with graphite electrodes for electrolysis is that some bubbles form inbetween sheets and manage to mechanically exfoliate flakes.