Why are dissolved oxygen readings off with a do-it-yourself sensor?

[Ronald.C]

I’ve been trying to measure dissolved oxygen in my aquarium water using a homemade electrochemical sensor with a graphite electrode. My readings are consistently lower than the test kit I have, and I can’t figure out if the issue is my electrode surface area, the stability of my voltage, or something in my calibration solution.

[ScarlettFA]

Fouling on the graphite surface is a real everyday issue in an aquarium. Biofilm, organics and minerals can coat the electrode and shrink the active area, which makes the current response look lower than a test kit for dissolved oxygen. Try cleaning or reconditioning the surface: gently polish the electrode with fine abrasive paper or a clean eraser, rinse with deionized water, and dry. Then compare readings in an air saturated sample at the same temperature as the tank. Also make sure there’s some mixing so you’re measuring the bulk water rather than a stubborn boundary layer; a still layer can understate the true DO. If it improves after cleaning, that points to surface fouling.

[Alexander79]

Voltage stability and reference drift matter a lot in DIY setups. If the reference electrode or the wiring is wandering, the polarization of the graphite will push the reading down, and you’ll see inconsistent lows. Check that the reference is clean and stable, use short, stout leads, and minimize any stray currents. If you can, test with a solid, stable bias source or a simple potentiostat to see if the signal stays put. If the reading still lags only for your DO sample but not in a simple buffer, that’s a hint it’s the electronics or the electrode conditioning rather than the chemistry.

[RileyFJ]

Calibration solution and temperature make big swings too. If the solution you used to calibrate is old or contaminated, the baseline drifts and your slope is off. Also DO solubility depends heavily on temperature, so calibrating at room temp and measuring aquarium water at 28 C will give skewed values. Use fresh calibration solution and calibrate at the same temperature as the tank, or at least measure and apply a temperature correction. I once skipped warming up the calibration water and the numbers looked way off until I did that.

[PaisleyBW]

Maybe the problem isn’t the sensor at all. I’ve chased similar mismatches and sometimes the test kit is giving a different endpoint than what your electrode expects because of different reference points or sample handling. Are you testing in the same water and at the same temperature as the kit’s reading? A tiny change in mixing, light exposure or sample volume can flip the numbers. It’s annoying, but I’ve found comparing a quick sample side by side helps decide if the issue is sensor drift or just measurement context.