I'm having problems with interference. I've read a lot of the prior posts on interference but not quite finding my situation.
I have an Amphenol temp/humidity sensor on I2C. If it's in one room the unit and code work great. If I move the unit exactly as is, nothing different, the values I get from the sensor chop around.
I'm hoping you can see the attached image. In the graph, the first part is room A and you can see in the graph when I move it to room B. In room A there are strong lights, A/C and such. Is this enough to cause such problems?
AC blows in cool air, with usually near 100% relative humidity. However this cold air quickly warms up as it mixes with the room's air, RH drops, and the end result is the typical dry AC air. That may very well explain your variation there and then.
RH is relative - take a room with a constant amount of water vapour in the air, and when you drop the temperature of that room the RH goes up, when you rise it the RH goes down, without adding to or removing from the air any moisture.
Nce graph! How did you get that?
It is good scientific practice if one suspects a sensor to malfunction, to do a measurement with another sensor, preferably from another manufacturer. Another rule of thumb is to measure in room A, move to room B, measure in B, go back to room A and measure A again. This will not rule out interference, but it will rule out malfunction of the sensor.
is your Arduino battery operated?
Air from refrigerating air coolers is dry because the water vapor in it condenses when it hits the cold evaporator coil then drains into a catch pan, look under your car after stopping on a hot humid day.
Thanks all. I never thought A/C would pose that much of a problem.
Thinking I'll create some sort of shield, even sticking it in a cup or something. The room temp and humidity is fairly consistent but the comments were correct; it is turning on and creating a dynamic I hadn't thought of.
It's going to be hard to stabilise your readings that way - and the question is really, should you, as it appears the sensor is working correctly and it's really the temperature/humidity that is changing so much.
The second half of the graph looks as expected. If you take a room with a constant volume and constant amount of water (kg of water) in the air - and subject it through the typical temperature changes that happen through a day, it looks like the second half of the graph. RH rises due to air being cooled during the morning hours (midnight to say 11am) and then drops as the air heats up (RH lowers because hotter air can hold more moisture than cooler air).
Whatever is happening in the beginning is pretty odd. I doubt that the actual room RH changes that much in such a short time. Is there anything heating/cooling/adding moisture to the sensor? Usually those RH sensors measure the voltage across a capacitor made from some silicon material that changes when exposed to moisture. Maybe there is some weird static causing the sensor to be charged.
I did confirm over the weekend that by putting the sensor in a paper towel tube (no, not something I really expect to do long term) that the reading stabilize i.e. ruling out issues with the circuitry. Now I need/want to find a way to protect the device from the AC and other air currents but yet get a somewhat accurate reading.
willowdog96:
Now I need/want to find a way to protect the device from the AC and other air currents but yet get a somewhat accurate reading.
Are you sure the readings you get are actually inaccurate? I think the readings are accurate, and your sensor is doing its job well. That the measurement results are not as constant as you would like to see doesn't make them inaccurate.
What you're trying it do is slow down the changes of air temperature and humidity, and that way levelling off the readings. This way you may get a nice and straight line - but probably a LESS accurate reading overall, as the sensor reacts much slower to its environment.