When Otto Wont

At some point, fixing an old autopilot just isn't worth the bother. Heres how to sort through the repair and replacement options.

Among all the things that can break in an airplane, autopilots are the most perplexing and potentially expensive, short of the engine.For years, they labor without complaint, dutifully holding heading and minding altitude. Then all of a sudden, the pitch control goes a little crazy or the needle tracking gets sloppy. Maybe its just a corroded connection or a tired servo. Then again, maybe not. Autopilots have a deserved reputation for being difficult and expensive to repair. At some point, you have to ask yourself if the box is worth fussing with one more time or replacing.

The truth is that late model autopilots-the Bendix/King KAPs and KFC or the Century 41s-are almost always worth repairing because replacing them is expensive and you wont get additional reliability. With older models-Pipers AutoControl series and so on-that may not be the case.

Many autopilots have characteristic failure modes. If your hangar mates Century 41 had a trim switch failure and you have the same vintage autopilot, don’t be surprised to have the same problem.Heres a short list of the bad things that can happen to popular autopilots and what to do when its your turn, plus a primer on autopilot basics.

How They Work
In general, there are two types of autopilots: Attitude autopilots and rate-based autopilots. Attitude APs use an attitude gyro with electric pick offs for roll and pitch sensing. Obviously, the attitude gyro is the critical component.

How is your attitude gyro driven? Is it electric or vacuum? Both types are used. Consider your plight, then, if youre in IMC with autopilot engaged and the vacuum pump dies.

If the system has heading track mode-as most do-the heading command signal from the directional gyro or HSI directs the autopilot to steer the aircraft to a specified heading.

Navigation tracking information is generated by a common nav source, such as a VOR/LOC receiver and or GPS or loran receivers output to the autopilot computer, which contains the electronic circuitry necessary to interpret autopilot command signals.

These command signals are sent as control voltages to the servos, which actually move the control surfaces.

Attitude-driven systems have a rep for smooth and precise rides but can be expensive to maintain because of high component costs and the fact that they have more parts. The bill will be even higher if you have a flight director.S-Tec has championed and done we’ll in the market with the rate-based autopilot design, in which a turn coordinator is responsible for roll outputs. An internal rate gyro sends a signal to the flight computer providing standard rate turn command for the roll axis.

The DG/H S I is responsible for heading track, just as in the attitude system. An absolute pressure transducer plumbed into the aircrafts static system handles the pitch axis and/or altitude hold function, sending a signal through the flight computer and then to the servo. Some rate-based autopilots have just pitch control, others have both pitch and altitude hold.

In general, field experience suggests a turn coordinator will outlast an artificial horizon, thus making a rate-based system cheaper to maintain and to install.

Some argue that a rate-based system doesnt offer as smooth a ride, since the turn coordinator bounces in turbulence. This is debatable and may have to do with the condition of the instrument.

A rate-based system probably offers more redundancy, since a vacuum failure has no effect on the electric turn coordinator operation. On the other hand, if the TC crumps, youre still toast.

And When They don’t
As with most aircraft systems, lack of use can promote component and overall system failures. Take for example an early model Piper Cherokee with even the most basic wing leveler autopilot.

We see the following scenario often: A customer purchases an airplane and advises that the autopilot is inoperative. After a few qualifying questions, we determine that the aircraft has been idle for a long period. Now that its back in the air, the autopilot wont even engage. Would it work if it had been used regularly? Perhaps. Advice: If you have an autopilot, use it.

Misuse and abuse is another occasional cause of autopilot failures. Landing the aircraft with the altitude hold engaged on a regular basis is definitely hard on the pitch servo and trust me, we have seen this happen more than once. The same applies for forcing the airplane into rolls with heading or wing leveling engaged.

Your AP no doubt came with an operators manual, which is required in a POH supplement. Review it now and again, especially with regard to limitations and ground-test modes. Ground test is there for a reason and may explain anomalies.

It will also explain what many owners consistently don’t know: Which instruments on the panel are tied into autopilot operation and which arent. This knowledge is critical to understanding how your system may behave if it fails and it could save your butt in an emergency.

Characteristic Failures
Heres a short list of bad things that can happen to popular autopilots and what to expect when its your turn. Attitude Driven: All of the systems manufactured by Century Flight Systems (and Piper Autocontrol) are vacuum-driven attitude systems. System components are almost always identical but model numbers are slightly different.

Usually, the only differences between Century and Autocontrol components relate to STC applicability. Basically, its a numbers game but its important and sometimes sticky when trying to replace older system components.

It might not be as easy as you think to swap an Autocontrol component with a component stamped with a Century data tag, due to STC applicability.

The single most common failure of a vacuum-driven system is related to attitude gyro performance. A sluggish gyro will cause the AP to command correspondingly sluggish pitch and roll control and often, these gyros don’t fail at once but over a period of time.

Pay particular attention to the attitude gyro at engine start. Does it take longer than used to erect? Is it sluggish rolling into and out of turns? Its not unusual for an attitude gyro to show these symptoms intermittently before failing entirely.

Replacement with an overhauled or new gyro is often the first step. If there’s doubt, try a loaner before spending money chasing other components or wiring.

Older systems such as the Century 21 and 41, for example, have complex wiring harnesses that connect to and from the gyros, servos and flight computers. With age, these get stressed and chaffed and create intermittent failures.

The harness assemblies arent cheap to replace and finding the problem-if it is the wiring-is time consuming. Recent Century harnesses seem more rugged. Some of the more advanced Century systems utilize multiple computer/amplifiers, which can lead to mode failures in automatic pitch trim operation and nav tracking/glideslope.

Trim problems: If the airplane has electric pitch trim, the control-wheel mounted switch is connected to the autopilot through a harness assembly that can be time consuming and expensive to fix. It may be better to replace the harness assembly when replacing the switch, just to head off a wiring failure later.

And the electric trim system, with its servo and accessories, is often a major portion of the autopilot system. Servos arent a common failure item, compared to gyros, but they don’t last forever.

An older servo may show signs of failure by sluggish control response, causing wing rock or oscillation and possibly blowing through the selected heading while in heading track mode.

Some shops say servos should be routinely overhauled ahead of failure but I think this is a bad idea, and costly to boot. You could sink more money into the servos than the system is worth. I would be cautious of a shop that recommends this approach when all seems to function satisfactorily. Servo failures, while not necessarily rare, arent the most common fault, either.

Rate Based
Rate-based systems from S-Tec seem reliable and often easier to troubleshoot because of their simplicity. Theyre more likely to fail hard, not even engaging if there’s a problem.

When the flight computer senses the proper output voltage from the spooled-up turn coordinator, the two handshake and the system is ready for use. A failing turn coordinator rate gyro will often cause intermittence thats apparent right at start-up.

Voltage fluctuations may cause the system to engage some of the time but not come up at other times. As with the attitude gyro, replace the turn coordinator with a known healthy unit or even a loaner to check the problem.

A check of the control cable rigging tension is sometimes worth considering. Seasonal temperature changes can slacken the cable and cause poor autopilot operation. When all the electronics check out-including the servos-control tension specs are the next thing to check.

When to Say When
In older airplanes, you can reach a point of diminishing returns with autopilot repairs. Say you have an older Century two-axis system in your Piper Arrow or Dakota.

Your shop troubleshoots the system and recommends that you change the attitude gyro, which is original equipment dating to 1979. They also found that one of the servos is fried and needs overhaul, plus there’s intermittence in the computer harness.

Sum total? It could $2000 to $3000 for a system thats out of date. Given how S-Tec has expanded the lower price segment of the autopilot market, you should shop a new one before okaying repairs costing that much.

We had a customer with a clean 1980s vintage Archer equipped with a Century 21. The nav coupling was inoperative and roll was useless. We checked through the wiring harnesses and reworked some of the connectors as a preventative. But the system continued to fail every third trip.

What to do? Shotgun each major system component? Not wise and certainly not cost effective. He replaced the Century with a new S-Tec model.

An even bigger headache is the Cessna 200 and 300 series Navomatics found in older single-engine Cessnas. Servos are often tired and critical components needed for system repair are expensive. Increasingly, these systems are becoming nickle-dime money magnets for their owners. Unless its an inexpensive fix, say under $800 or so, consider replacement. Keepers

If you own an airplane with a 1980s vintage Bendix/King system thats giving trouble, you may have to bite the bullet and keep fixing it.

Just about any Bendix/King KAP/KFC series system is a keeper. These are often original manufacture supplied in high-end airplanes such as Bonanzas and Barons, later Mooneys and Malibus.

These systems are quite expensive to replace and you arent likely to get any added reliability from a newer version of the same thing. These systems tend to be feature rich, with flight directors, auto trim and glideslope coupling, stuff that doesnt come without a maintenance obligation.

All of the S-Tec APs, in my view, are worth repairing and/or upgrading. S-Tec has adopted a building block philosophy, making it easy to add altitude hold and so forth.

S-Tec has proven that system upkeep and maintenance is cost effective and most shops don’t complain about working with their stuff. Tech support is excellent and component exchange supply is abundant. The higher-end Century equipment, the Century IV, for instance, is a we’ll performing system packed with features. Upkeep of this system is logical and not overly expensive. The same advice applies to the Century 2000, a relatively state-of-the-art AP.

Id put the higher-end Cessna equipment found in medium and cabin class twins in the same good-to-keep category, although we find many customers upgrading these autopilots to more modern systems simply because the airframes have appreciated enough to warrant it.

Wrap Up
When making a decision to repair your older autopilot and related components, note the spirit of your shop. If they seem hesitant to get involved with the troubleshooting process, find out why. You may be able to talk them into anything but their reluctance may stem from the chronic problems they know youre in for.

Discuss upgrade alternatives and have them quote a new system with similar features as your broken system. You may still decide to fix the old autopilot but at least you’ll be able to attach some value to fix versus replace.

If the shop suggests an overhaul everything approach, ask questions about the bottom line and whether or not theyre overhauling serviceable components because theyre out of ideas on how to fix the old ones.

Any shop should offer to fly the aircraft and evaluate its performance, which gives you a chance to make sure youre operating the AP correctly.

And last, be precise when you explain your systems problem to the shop. This is the first step in making the right autopilot repair/replacement decisions.

Also With This Article
Click here to view “Common Failure Modes.”


-by Larry Anglisano

Larry Anglisano is an Aviation Consumer avionics editor and a consultant for Exxel Avionics in Hartford, Connecticut.