PlasticPilot.net

FGC recently dropped a question my “questions box” about my experience of the Piper Saratoga, avionics and possible update to a glass-cockpit. I will answer it later, before I wanted to react to FGC’s remark on auto-pilots.

“I was wondering what is your experience with it in the Saratoga and if you have had any change to hear about autopilot upgrades for the Saratogas from the KFC150. I have previous experience with the STEC 55 in Cirrus aircraft (about 20 HRS) and it is a much better experience.”

What I like in this remark is its honesty, and the fact that I feel exactly… opposite ! I flew a lot with the KFC150, and it’s by far my favorite autopilot - the only thing missing is an altitude preselect. I also used the KAP140 that equips G1000 DA40. My STEC experience is limited, but I don’t really like this autopilot. This is why FGC’s remark sounds interesting. Why such a difference in opinion ? Here are a few ideas…

A good friend, ready to kill you
Autopilots do great jobs at keeping heading, follow a radial, maintaining altitude or simply keeping wings level. They even do it much better than humans, because they feel any mini deviation before it becomes visible on instruments, and are fully focused on the task whereas pilots have many other things to manage.

Turning the autopilot on certainly reduces the workload, but they do only what they are programmed for. Set a heading and altitude that will fly into terrain, and the autopilot will trustfully bring you there !

Something I dislike on the STEC and the KAP140 is the vertical speed mode. The pilot defines a rate of climb, and the autopilot adapts pitch to maintain it. If the rate decreases, the autopilot increases the pitch angle. With no consideration of airspeed. Engage this mode and let the plane climb high enough, it will result in a stall !

Know it - Control it - Profit from it
Different autopilots have different philosophies and interfaces. Typically, with the KFC150 it is possible to select the modes (HDG, NAV, ALT, …), then visualize how the flight director before engaging the autopilot. Doing so, it is possible to avoid engaging the autopilot if the setting is wrong.

This is not possible with the KAP140, which must first be engaged and then programmed. Another difference is that each button of the KFC150 has one and only one function. The buttons of the STEC and KAP140 have various functions, some acting as mode togglers.

I’m not saying that one is better than the other. What is important is to know how your autopilot is working, and how to make it do what you expect.

I was once flying as safety-pilot with a friend of mine. On a relatively complex IFR approach, my friend did not programmed the autopilot correctly, and got surprised by what went on. His reaction was to disengage the autopilot and fly the approach manually. Detecting the problem was good, and disconnecting the autopilot was a correct decision, but it lead to an unexpected increase of workload. It would have been better to use it correctly.

It is also of paramount importance to know how to disconnect the autopilot. If a malfunction leads to a massive trim action by the autopilot, you’ll have to disengage it to regain control. AP disconnect on the yoke, AP button on the box itself, circuit breaker, and trim servo circuit breaker are all possible ways to neutralize a misbehaving autopilot.

No Macho Attitude
One of the permanent argument in the pilot community is “manually flying vs. autopilot use”. On one hand, the tenants of manual flying say that pilots shall be able to fly with no help from the autopilot because it can fail. On the other hand, autopilot fans say that the workload reduction makes possible for the pilot to focus on higher level tasks while the autopilot is flying.

This debate sometimes sounds like a war of religion, and as frequently the solution is the way in the middle. While it is true that all pilots should be able to fly manually, autopilots allow for more safety and precision, if you know how to use it. I personally fly manually on a regular basis, but most of my IFR time is spent with autopilot engaged.

If I had once to fly a complex approach in IMC with no autopilot, I would certainly request vectors from ATC, to reduce the workload. Call-me coward, but I will not start an IFR flight if the autopilot is not working…