Those values are not so different, average 909 with variance +/- 1.2%.
All hardware has a tolerance. For soil moisture I would be very happy with this tolerance.
yeah since this post it dropped by 100 and few measurements later it went up by 100 againā¦
itās all funkyā¦I removed the transistor, Iām sending the VCC from the pin on the esp. and ground from ground and I know the measurements of the 3.3V with this sensor so I mapped it to thatā¦ Iāll see how that goesā¦
this way there is little to no voltage difference when not in use which means the sensor will live longer and when I need I already send the voltage neededā¦ again Iāll see how that worksā¦ thanks anyways for all the help you gave with this. 
really appreciated!
Sounds like your transistor circuit is faultyā¦ or at 50ms between power up and reading you are switching on and off too fast for the transistor?. I have had good results with basic 2n2222 for switching higher power items, and they are high speed, but I also donāt worry about 24/7 power to sensors. I just leave the sensor powered and cover all connections in dielectric grease (if outside or in high humidity areas). 24/7 power doesnāt kill the sensor, corrosion from DC & moisture does.
Over sourcing current from the GPIO pin may or may not hurt your ESP over time, but againā¦ who knows for sure until it happensā¦ and the MCUās are cheapānuff to replace 
Marking this topic as solvedā¦ but feel free to add to it if you have more info.
1 Like