Autopilot installation guidelines

Although it was more than a decade ago, I remember the passage very well. I was on a short-handed delivery from Bermuda back to New England after acting as navigator for a Marion-Bermuda Race. I was helping the skipper return the boat and was looking forward to a leisurely passage. Somehow, it’s always the passages you count on to be leisurely that are anything but.

Within hours of leaving St. David’s light in our wake, the boat inexplicably rounded up. After disengaging the autopilot, my shipmate took the wheel while I did some investigating with a flashlight. The autopilot drive motor – a chain-and-sprocket affair located in the binnacle under the wheel – had wrenched itself free from its plywood base, breaking one of its nylon gears in the process. The remainder of the voyage was plagued by a number of other gear failures as well as wave after wave of squalls and heavy lightning storms, all of which had to be endured at the helm because we could not repair the autopilot drive with the onboard spares.

Where autopilots are concerned, proper installation is as critical as selection, sometimes more so. Alas, a high-dollar, full-function system installed poorly is worthless compared with an economy system installed properly.

Before beginning the installation, or contracting a technician, take a moment to lay out the system, mentally first and then on paper. In this case, the axiom of the five P’s applies: proper planning prevents poor performance. Try to visualize where each component will fit, where cables will run and what storage space will be lost. Take into account serviceability and access for repair operations. How difficult will it be to access each component, especially the drive mechanism? Will the drive or tiller arm (the device that connects the drive ram to the rudder) interfere with sails or other gear stored in this space? If so, a partition must be installed.

If the autopilot uses a separate rudder angle transducer, then this must be protected as well (rotary rudder angle transducers are considerably more delicate than the linear variety). A dock line, lifejacket or drawstring snagged on this very delicate device could cause hundreds of dollars of damage, as well as an autopilot malfunction, with a single pull.

Drive units need solid base

The autopilot component that I’ve most often seen installed inadequately or incorrectly is the drive unit. To understand the loads imparted to this area of the installation, carry out the following test. On a tiller-steered boat, steer as you would normally, holding the tiller near its forwardmost end. Then choke up to within 12 or 18 inches of the rudder. It’s almost impossible to control the rudder, especially in a beam or following sea. This is a crude but telling approximation of the loads a linear drive experiences for hundreds or thousands of hours using the average-length tiller arm. A chain-and-sprocket or geared drive may face even greater torque.

Accordingly, these drive units must be secured to the hull solidly. Only through-bolts should be used when securing the drive mechanism to a stringer, bulkhead or well glassed-in shelf, and backing plates are a must. Do not use tapping screws or lag bolts.

Where linear rams are concerned, be they hydraulic or electric, the tiller arm may be attached to the rudderstock at any location around the shaft as long as it fully engages the key. It must, however, be perpendicular to the tiller arm when the rudder is amidships, in order to achieve the proper geometry and impartation of force to the rudderstock. Some quadrant manufacturers prohibit piggybacking a linear autopilot drive onto the quadrant, and this makes good sense. Although it’s more time consuming and costly to install, using an independent tiller arm affords a greater degree of redundancy.

Separate tiller arms

In the most popular types of steering systems for small sailing vessels, cable over sheave, jacketed cable, or a geared tie-rod system (Whitlock or Edson), if the cable parts, the quadrant slips or some other component gives up, the boat can be steered with the autopilot if a separate tiller arm is used. Remember, bronze and aluminum components are galvanically incompatible. Rudderstocks may be bronze, stainless or sometimes aluminum (particularly on an aluminum vessel); quadrants, and tiller arms may be aluminum or bronze.

Ensure that metals that are unfriendly to each other, particularly aluminum and bronze, are not installed so they remain in direct contact. Water dripping off a bronze tiller arm onto an aluminum quadrant or hull could cause severe galvanic corrosion over the course of months or years.

A final word on drive-unit installation: It is critical that the drive unit, if linear, have more travel than the tiller arm. The drive unit should never act as the rudderstop or shorten the travel of the rudder.

If there is to be only one display, it is probably best to install it near the helm so it can be adjusted, monitored and disengaged easily. If installing multiple stations, or a hand-held remote, the main unit should go at the nav station and the remote at the helm (I have installed repeaters in the skipper’s bunk on many a cruising vessel). Most remotes use a multiprong plug (with the trend toward wireless accessories, some remotes are now unencumbered by this electronic tether), and thus more than one location can be chosen for remote operation.

The electrical consumption of the drive unit can be prodigious, especially in a seaway, and thus the desired voltage drop of no more than 3 percent must be achieved. The calculation for this formula must include a “there and back” figure for the electrical supply cable. If, for instance, the drive unit is 20 feet from the main power distribution panel, the voltage drop formula would use 40 feet of cable for the overall run.

If in doubt, consult the ABYC Standards and Technical Information Reports for Small Craft (available at or by calling 410-956-1050), or one of the many marine electrical books that contain reprints of the voltage drop calculation tables. In addition to performance and efficiency issues, the life of an electric motor may be noticeably shortened by low-voltage scenario, and undersized cabling will surely exacerbate this condition. Make the connection between the drive unit and power supply via an insulated marine-rated terminal block to facilitate the difference in wire size between the drive unit and vessel side of the cable run, as well as for ease of serviceability and replacement.

When making electrical connections, include service and drip loops in each wiring run, and heat shrink connectors offer the greatest degree of strain relief and water resistance. The over-current protection for the entire autopilot system should be a circuit breaker as opposed to a fuse, and it should be easily accessible and well labeled for the purposes of resetting. It also should be dedicated to the autopilot, sharing no other loads.

Most installation manuals are fairly specific about where and how to mount the fluxgate compass. Needless to say, the farther this piece of gear is mounted from large iron masses – such as the engine and anchor/rode, or electric motors, pumps and high-current cabling – the better. Post placards in lockers and storage spaces adjacent to the fluxgate compass, prohibiting the storage of ferrous objects (one can of beans can ruin an autopilot’s entire day, and yours). Temporarily install the fluxgate until sea trials are complete. This will enable you to ensure it’s well placed and not affected by any of the aforementioned objects.

Swinging the compass is a critical part of the installation. Follow the manufacturer’s instructions meticulously. When doing this, it is best to energize all loads that will normally be on when sailing, both day and night. Electric current sets up a magnetic field and sometimes will affect performance and accuracy of the fluxgate, or any other, compass.

By Ocean Navigator