I’m still getting I2C errors logged, and the quality of each flight seems to depend on how many of these I get and when. They all occur during the warm-up code, and could be anywhere from zero to five or six. If they occur in the first few seconds of warm up, the impact on the quality of flight is small – later in the sequence and it becomes larger. That’s due to the fact the 20s warmup is primarily there for filling up the Butterworth filter history – earlier errors are less significant.
I do wonder whether it’s the altimeter also on the same I2C bus that may be causing the problem. My code doesn’t bother initializing it, so next step is to add code just for that. Fingers crossed.
I added the few lines of code to just initialize the barometer, but that made no difference whatsoever to the I2C errors so for the moment, I’m out of ideas.