Safe Aviator: Minding Autopilot and Pitch Trim Failures

A friend with a Mooney has attention-getting stories of elevator pitch trim failures while at low altitude and low airspeed. When we sift through the wreck reports while researching our Smart Plane Shopper feature, we find plenty of drop-short wrecks where the pilot manages to successfully get the airplane out of the low clouds only to lose it while maneuvering to land close to the runway. I wonder how many are autopilot failures—or the pilot’s failure to spot them at the last second. The failure in the Mooney with its automatic stabilizer trim system was a startling discovery after disconnecting the Honeywell KFC225 autopilot on short final after a coupled LPV approach, the out-of-trim airplane pitching aggressively and needing lots of counteracting control input to wrestle it into the landing flare. It’s a setup for a prang.

False Sense of Security

Some failures are pretty subtle until they aren’t. All seems well with the autopilot’s altitude hold/pitch mode engaged, with the system’s approach mode following the localizer and glideslope on down as the pilot adjusts the power. Add speed reduction, flaps, and landing gear deployment, followed by re-trimming the plane’s pitch axis.

In the digital KFC225, automatic pitch trim only works when the trim servo motor runs—at its full motor speed. When it doesn’t and the pilot doesn’t manually trim with the trim wheel, disengaging the autopilot usually means the pilot gets an airplane grossly out of trim and worse, with decaying airspeed. I’d call some failures an emergency.

In some airplanes, the emergency procedure for runaway or frozen pitch trim calls for both pilots to push or pull to counter the control forces. Trim failures in lighter planes with lower control forces might have less drama, with autopilot and trim servo slip clutches. And while you’ll likely muscle it, you can override the clutches to gain and maintain control.

For trim failures, there are warning signs and a call for vigilance in watching automatic trim systems like a hawk, and always be aware of how the plane is trimmed. When practical, an old trick is to rest a leg near the trim wheel so you know when it’s running or not. Better are integrated flight displays that show the pitch trim status (and flap setting) onscreen.

Preflight It Right

No matter the trim system, it starts with a good preflight—often easy to rush through. It’s the same theory for conventional elevator and servo trim tabs or with a stabilator and anti-servo trim tabs, so start directly at the trim tab and inspect the hardware and any control slop. Make note of its position on the airframe and then look at the trim control and annunciators in the cockpit. Do they match? Manually run the trim through its entire range of motion with the trim wheel, feeling for stiffness and binding. With good feel, you might spot a fray in the cable. If equipped, run the electric trim in both directions, listening to the servo motor and noting its speed and that the trim command buttons don’t stick. Ask your shop to carefully inspect the cabling and hardware at every annual inspection. It’s a maintenance chore that even some shops overlook.

Study the system in your plane and understand how it’s supposed to work. As one example, there might be a limit switch, one that opens the circuit powering the pitch trim servo and shuts it off when system components reach a predetermined full-up or full-down position as backstop. Know exactly where the autopilot and trim circuit breakers are, testing them regularly making sure they indeed pull power off the systems. Fresh out of avionics or maintenance work that had hands in the electrical bus? I’ve seen techs connect autopilots and trim systems to mislabeled circuit breakers. You might put circuit breaker caps on them for easy recognition.

But perhaps the most important preflight check you can make is running the autopilot’s before flight check. The test sequence varies widely by system make and airplane interface, but in the case of the KFC225 in the Mooney (and other BendixKing systems including the KAP140), the autopilot programmer/computer runs a self-test at each power up, sending signals to the servos for a clutch engage and disengage test. On the Honeywell system, the preflight test sequence is indicated by “PFT” with an increasing number for the sequence steps. Successful completion of self-test is identified by all display segments on the programmer being illuminated, and the Flight Director command bars brought into view, followed by the disconnect tone sounding. The S-TEC 55X system, as another example, has pages of preflight test procedures including pilot input moving the controls (overriding) while the system is engaged, plus pitch trim tests. Follow the preflight, normal, and abnormal procedures in the flight manual for the system in your airplane.

Modern Systems Aren’t Immune

I went to the Emergency Procedures section in the Flight Manual Supplement for Garmin’s GFC 500 autopilot and quickly found the procedures for pitch trim failures. While automatic pitch trim is an option on some systems (which makes it a trim-prompting system, where the autopilot prompts the pilot to manually adjust the trim through airframe configuration changes), many have full-up automatic trim. For an autopilot/pitch trim runaway situation and where the airplane “deviates from the planned flight path,” the procedure is to firmly grip the control wheel while pressing and holding the autopilot disconnect/trim interrupt button. The caution subhead warns of high elevator control forces, and to regain and maintain aircraft control. While still holding the disconnect button in, pull the circuit breakers.

The flight manual also warns not to overpower the autopilot because the trim will operate in the direction opposing the overpower force, which will result in large out-of-trim forces. And of course, don’t try to re-engage a system with a known failure. Garmin’s GFC 500 and GFC 600 systems have good onscreen and audible status annunciation. In a failure, the autopilot disconnect may be accompanied by a red AFCS in the autopilot status box, indicating the Automatic Flight Control System has failed. The flight director will not be available and the autopilot cannot be re-engaged when the annunciation is present. Ask your shop to demonstrate a newly installed autopilot, and train for failures on systems that are new to you, limiting operations to VFR until you grasp it.

It’s Also Why Autopilot Repairs Are Pricey

There are still plenty of vintage analog autopilots in service and many have problems. My sense is with all of the Garmin GFC autopilots that shops are installing, rare is the shop that’s equipped with the right equipment and troubleshooting experience to fix legacy autopilots for a minimum downtime and cost that make sense to keep the old system in the first place. Swap out a couple of servos and add some troubleshooting labor, and many legacy autopilot repairs flirt with $10,000 or more—especially when mechanical gyro instruments are involved. For many, it’s worth a quote for a replacement to at least keep the repair costs on the rails. Autopilot work is what surprisingly big invoices are made of.

What’s alarming to me about the failure in this KFC225-equipped Mooney is that the autopilot on multiple occasions successfully passed the preflight sequence, yet the entire trim system was inoperative in flight, perhaps after warmup. But ultimately, it’s a hard failure of the servo motor—you couldn’t command the trim from the yoke switch and the autopilot wouldn’t run it. The first step in the troubleshooting sequence might include accessing the trim and pitch servos, while inspecting the cables—including cable tension. Then remove both the trim servo and pitch servo for bench test.

Many shops will send failed servos off for third-party repair. One good source for Honeywell/BendixKing servo repair and exchange is Mid-Continent Instruments and Avionics, which has bench repair and overhaul capabilities for a variety of autopilot components. Autopilot upkeep isn’t cheap. Swapping out a couple of servos even for vintage systems could cost well north of $5,000—a necessary chore for safety of flight.