search

Hiverize / FiPy

3 stars 11 forks source link

filter wrong measurements #6

Open DieDiren opened 5 years ago

ClemensGruber commented 5 years ago

I got some strange measurements from a DS18B20 today. This was one of the first datasets the node sends, so I fear this will be not a single event. All other sensors reported normel temperature values but one a temperature of around 60 °C:

2019-05-23 15_55_28-Bob-App - SeaMonkey

The best way would be to build an median of 3 to 5 sensor readings. But for the DS18B20 this would mean up to 5 seconds working time (and power consumption). So we should find a solution and not only a workaround.

I'm reading the DS18B20 with other nodes and C-/Arduino code without median or mean building on software side so we should have a look at the used lib. I remember a checksum of the DS18B20 but must have a look again.

ClemensGruber commented 5 years ago

I observed three wrong readings from 15 to 22 h on a 20 second upload intervall with 5 different DS18B20. One reading was wrong on sensor #3 and 2 wrong readings on sensor #2.