David Megginson, the blogger behind the Land and Hold Short blog, has written lengthy post about selecting an autopilot for his aircraft and then flying with the one he ultimately choose (the S-TEC System 20). David began his post by noting:
I’ve flown over 850 hours, about 120 of them in actual instrument conditions, and all by hand. I’m proud that I can do that, but at the same time, it’s tiring: after 8 hours bumping around in cumulus clouds and dodging storms on the Stormscope, I’m exhausted beyond anything I’d ever imagined. Even 4 hours in good VFR weather can be tiring, because of the constant attention needed to keep the plane straight, level, and on course.
In other words, it was time to think about getting an autopilot installed.
David opted for the S-TEC System 20, a rate-based autopilot built into a turn coordinator. Its also a single-axis autopilot — meaning that it will control roll and heading but not pitch, altitude or yaw. However and had David opted for the S-TEC 30 with altitude hold, he would have needed to spend a few extra thousand dollars more than what he had budgeted plus he would have needed to have extra maintenance work done on his existing avionics.
On his way home, David got to test the autopilot for a good two hours and he noted that up at the 9,500 feet level right above a broken layer of cumulus cloud, he was able to fly hands off with just a minimal adjustment every 5 minutes to the elevator trim to hold the altitude. On the other hand and after he had to descend below the cloud layer, the flight became the equivalent of a mechanical rodeo bull that forced him to trim every 30 seconds or so as his pitch and altitude gyrated in rising and falling air columns.
In other words and in smooth flying, an autopilot without an altitude hold is just fine but in turbulence, a pilot will wish that he or she had spent the few extra thousand dollars to buy the altitude hold. Nevertheless, David noted that the S-TEC 20 can be upgraded to include both altitude hold and electric trim.
David ended his post by noting that his new autopilot will be a big part of his flying now and will likely encourage him to fly further afield.
I have flown a single-axis autopilot and found the same thing: it worked much better than expected as long as the air is smooth, but is no real comparison to a 2-axis.