2

I wish to measure microvolt-level fluctuations in a small volume of electrolyte solution, currently saltwater. My existing setup works but has some strange data so I'm hoping someone with more experience can help me sort it out.

Equipment

I'm using an electrically-isolated voltmeter sensitive in the ~30 uV range at a 100 Hz sampling rate. The sensor wire (red) is immersed in the solution along with the BNC ground wire (also coming from the sensor) which is then split to connect with Earth ground.

enter image description here

I've validated this setup by placing an additional wire into the solution carrying a 1 Hz sine wave produced by my sound card, which is picked up by the sensor as expected:

enter image description here

Questions

  1. Is this circuit really doing what I expect it to? It seems to work but I don't know exactly why. Do I even need the Earth ground if everything is shielded?

  2. When I use distilled water I get roughly the same reading (!?) and can also induce the sine wave. I didn't think distilled water was at all conductive, so... how is this possible? I've tried to carefully clean all equipment, new tube, new leads, no touching the wires, etc., although I'm trusting Arrowhead that my water is actually distilled...

  3. When using saltwater, the voltage seems to increase slowly over time, even when EMF shielded (see chart.) I had expected it to be more constant on average or go down--- what might account for this behavior?

Thanks!

enter image description here

drc
  • 123
  • 5
  • 3
    Water has a huge permittivity (relative permittivity aka dielectric constant in the range of 80) so it is easy to get a few uV of capacitive coupling of AC signals. Do you have any way to independently measure the source resistance of the test electrodes? – Spehro Pefhany Mar 31 '22 at 23:34
  • Capacitive coupling would explain the voltage increase over time you mean? Do you know why I would get a reading in distilled water? Re: electrodes, no, I couldn't seem to find a good "standard" BNC-fitted electrode that wasn't designed for pH or ORP testing, so I'm just using BNC split into two bare wires... which I imagine react poorly with saltwater. If I knew how to measure source resistance I could try it. Fortunately I don't need an absolute measurement, I'm just looking at relative variance. – drc Apr 01 '22 at 00:01
  • 1
    No, it would explain the distilled water result. Increase in leakage current as the board warms could explain the increase over time of the salt water, or maybe it's an actual effect. Without numbers it is difficult to guess. You could try measuring resistance with a multimeter however that will immediately cause electrochemical effects with most multimeters since they use volts, not mV. – Spehro Pefhany Apr 01 '22 at 00:05
  • 1
    P.S. My ordinary Fluke meter I mainly use for mains measurements (because it has proper protection) has a 200nS range with resolution 0.1nS. – Spehro Pefhany Apr 01 '22 at 00:12
  • Thanks for your replies. Would capacitive coupling also explain how I can push an audio signal through distilled water? – drc Apr 01 '22 at 00:19
  • 1
    That's the only thing it explains. – Spehro Pefhany Apr 01 '22 at 00:22
  • 1
    If you want "measure" ac resistance of "electrolyte" ... see docs AD5934 – Antonio51 Apr 01 '22 at 06:13
  • If it's capacitive coupling, it should not occur when the signal is DC, correct? – drc Apr 01 '22 at 16:46
  • If the purpose of the experiment is to measure (DC) contact "potentials", it is certainly necessary to know the type of chemical reaction that occurs in the case considered ... which may not be simple, if the electrodes are not "noble". The temperature will certainly play a "big" role. example: https://www.osti.gov/servlets/purl/1337148 – Antonio51 Apr 01 '22 at 17:08
  • This would be only for the purpose of validating whether the current is indeed being carried through the distilled water via cap. coupling. If I instead pushed various levels of DC current through, it should not affect the measurement value because there is no electrolyte present to carry the charge and no AC to provide coupling. – drc Apr 01 '22 at 17:58
  • Hum... I think that DC (although very "low") will always have an effect because of dissolved "impurities" (reactions with water), present also in "distilled" water. – Antonio51 Apr 02 '22 at 08:07

1 Answers1

1

I have a suggestion to debugging your setup methodically.

  1. Determine the source resistance of the configurations you need to measure. By direct measurement if at all possible.

  2. Procure a resistor or resistors of equivalent value. If you are in North America that's a next-day thing.

  3. Test with the resistor(s) in place of the electrodes and compare results.

Given the relatively large voltages from electrochemical reactions, I imagine you are going to have to be very careful with the electrode composition and cleanliness (maybe identical platinum wires from the same piece), but that's just a guess.

I have certainly seen leakages that could explain what you are seeing due to subcontractors using unapproved procedures. That should be caught by the module manufacturer, but..

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842