Not sure if accuracy is the right term. Here is the scenario. I have two weather stations I have built. One is a Dev box and one is a Test Box. Both are configured exactly the same and use the same firmware. When I got them together a few months ago the Pressure readings from each BMP180 were identical.
The Test box was then moved outdoors for the last few months. Over time the outdoor BMP180 now reads .04 higher than the Dev box which has remained indoors. The outdoor BMP180 is inside a stevenson screen. I am wondering if the reason for the difference in readings could be dust that has settled on the sensor ??
If you mean 0.04 hPa, its probably just that the two sensors are off-kilter from each other. If you read the datasheet, you can find the absolute accuracy of the device:
gfbuyer:
When I got them together a few months ago the Pressure readings from each BMP180 were identical.
Identical? That would be remarkable. The noise of these sensors is roughly the same magnitude, a few hundredths of an hPa. But even average values can be off by a lot more. I have two BMP180s and two BMP280s and they all report a little differently. The average values of the four vary by something like ±0.40hPa. And they also drift a little relative to one another from hour to hour, day to day.
Here is a histogram that someone else created showing the output of two of their sensors. The x-axis is in units of Pa, so the peaks are about 0.50hPa apart.
While I did not graph the results, after a period of casual observation with the two running next to each other, yes the results returned were either right on or different by a hundreth.
I guess the issue I am seeking help with is related to the change from being very close to one another to the point of indistinguishable difference, to now a variance that is consistently .04.
I am wondering if perhaps the outdoor conditions have contributed to the change and asking for anyone else who may have experienced something similar.
I don't know the fundamental causes for long term drift of these sensors. Bosch states in their datasheet that it can be as much as 1.0 hPa.
Even with two sensors that are in good agreement there is no reason to assume that they are accurate. It would just be chance that they are that close. According to Bosch the typical absolute accuracy is -1.0 ± 1.0 hPa.
What I don't understand is how you are getting precision of 0.01 or 0.02 hPa with a sensor that has RMS noise greater than that. How are you reading this sensor? What mode? How often? How do you process the data?