In article ,
"Matt Barrow" wrote:

Some AP's require you to disengage during moderate or stronger turbulence.
There's also a big difference between rate-based and attitude-based.

Is there such a thing as pure "rate-based" vs. "attitude-based" in modern
AP's any more? Back in the days of steam gauges, each state variable
involved another mechanical data pick-off and gobs of discrete circuitry,
so controlling one variable at a time made sense.

In a glass cockpit, I would think you would grab all the data that's
available and come up with a control strategy which mixes inputs.