I have connected my Arduino with an YT-38 to measure the humidity of a glass of water. I used the YT-38 with two different sensors (YL-69 and a not-named one from China).
The sensors are turned on two seconds before measuring, then they measure, then they are turned off until the next measurement (to prevent electrolysis). The next measurement takes place after 20 seconds, the whole measurement time was about 20 - 30 minutes (depending on the sensor, but this difference should not be relevant for this issue).
That sensor measures resistance, not "humidity". Humidity is the water vapour content of air.
Clearly the random one from China is a lot better - perhaps better electrodes or less noisy
circuitry or less noise-pickup from the wires? Perhaps there's a poor connection somewhere
and its a transient problem?
You could turn on the sensor 10ms before the measurement I suspect and reduce electrolysis a lot more.
Why is the China sensor better? Okay, it doesn’t measure peaks but it still measures different values (from beginning to the end of the whole measurement-test it’s a decreasing curve)? So the main question is why the measured values are a curve (for both sensors) - any ideas? The connections are quite simple, I would not know where I could have made a wrong / problematic connection... I’ll double check that.
EDIT: I reduced the time to 25 ms and the measurements are constant now (for both sensors). So it seems that this was the solution.