In a previous article on autopilot sizing we looked at rudder movement, flow rates and pump sizes. The drive system of an autopilot is matched to the vessel’s steering system. This is how we achieve appropriate response for hull design.
This month we will take a wider view of the autopilot system and talk compasses. Generally, the electronic part of the pilot consists of four major components: controller, compass, computer and rudder sensor. These components interact to produce steering commands to the drive. Although manufacturers differ in their configurations with virtual rudder feedback and computer combined into compass or drive motor housings, all autopilots share the same end goal, to steer the vessel reliably. In practice it is not uncommon for autopilots to fail. They have a very difficult job to do that is both complex and high torque. Moving a rudder is hard enough, but the pilot must also move the rudder to the correct angle for the correct amount of time and then remove that rudder. So autopilot failures range from dead patient to unstable performance. “S”-ing is the most common complaint, and this can be caused by oversteering or understeering. Manufacturers have worked hard to avoid the symptom “makes sudden turns,” but we still see this happen. To ensure a safe and reliable autopilot installation, do it by the book and understand what you are doing. Always instruct your customer on operation and potential dangers.
The compass To understand autopilot function, we start with the compass, the eyes of the system. There are several heading sensors available; the most common is the fluxgate compass. Not to be confused with the flux capacitor used for time travel, the fluxgate compass consists of coils of very fine copper wire wrapped around a ferric core hanging on a gimbal. When accelerometers are added, we call this sensor a rate sensor. When the lines of magnetism from the earth cut through the energized coil in the fluxgate, a small voltage change is produced across the coils, depending on the strength and angle of the earth’s magnetism. Once fed into a microprocessor and combined with the accelerometer data, a digital signal such as an NMEA 0183 data sentence or an NMEA 2000 ® PGN is output. The heading sensor is the heart of the autopilot system. Heading sensors come in a few flavors, like gyrocompasses, satellite compasses, fluxgate/rate sensors and others. This sensor tells the computer where the bow of the boat is facing. Remember, heading and course are not the same. In 0183 parlance, we call magnetic heading HDM (Heading Magnetic) and GPS course COG (Course Over Ground). COG is the direction the vessel is traveling, HDM is the direction the bow is pointed. To steer the vessel we need to know where the bow is pointed (heading). Heading is a data point that many vessels lack until an autopilot is installed. This can be a complex subject, since we reference both magnetic (north pole) and true (center of earth’s rotation). The magnetic and kinetic forces of the earth are beyond this discussion, but we do use both magnetic and true heading extensively. Basically, if you sense magnetism you must compensate for position relative to the North Pole. If you sense rotation, you must compensate for earth’s velocity.
Sensors and heading data The North Pole is to the left of true north, that is we use a westerly variation to compensate, but only from here. In Asia and Europe, things are different, of course. The velocity of the surface of the earth varies depending how far north you are on our spinning planet. We compensate for this with latitude and speed sensors. The Advanced Marine Electronics Installer (AMEI) class offered by NMEA covers heading sensors and heading data. Some devices such as a mandatory Class A AIS require True Heading data to be input. When it comes to consumer autopilots, we typically use a rate sensor supplied with the autopilot package to sense magnetic north. Whether we sense true or magnetic north, we use this sensor to keep the pointed end of the boat pointed the right direction. The heading sensor is also a requirement for radar chart overlay and MARPA. More on heading data in a later article. Another sensor used for autopilots is the rudder sensor. Of course, this tells the pilot where the rudder is. In a simple form, this is a potentiometer, like a light dimmer, that varies with rudder movement. Simrad uses a frequency rudder sensor that outputs a frequency signal which varies with rudder movement. Most manufacturers also offer virtual rudder feedback (VRF) where there is no rudder sensor at all. A VRF type system is typically used on outboard and I/O applications. They use black magic and software trickery to dead reckon the rudder position based on pump run time. Without a rudder sensor, there is no RAI, Rudder Angle Indicator, a desirable feature. A rudder sensor allows the pilot to steer more efficiently and should be installed when practical. Some applications still require a rudder sensor, such as a catamaran with outboards. Linear rudder senders are used for applications where a rotary unit is not feasible.
Sensor location is critical The rudder sensor requires near perfect geometry to function properly. Since steering systems vary greatly, there are many configurations that are used, but an even, linear response both port and starboard is desirable. This means that the physical mounting of the sensor is critical and can be challenging. Improperly mounted or configured rudder sensors may cause the pilot to steer differently to port and starboard, which is disastrous for pilot performance. A typical rotary rudder sensor should show a rectangle at midship and a parallelogram at all rudder positions, hard over to hard over. The AMEI class teaches autopilot installation techniques.
A basic autopilot includes a compass and a rudder sensor. When turned on and engaged, the pilot tries to keep the heading steady as she goes. When the boat wanders off course, a heading error is generated and then corrected by applying rudder opposite the heading error. Once the error is eliminated and the boat is back on course, counter rudder is applied to stop the turn. This most basic pilot function requires an accurate and responsive compass and rudder sensor. Calibration of the rudder sensor, setting limits and center are normally required. Compass calibration to compensate for deviations is also usually required. The physical planning of the installation can make all the difference. Perfect geometry on the rudder sensor, good compass location—these things matter. Poorly installed equipment may appear to work, but will not perform in adverse conditions.
Autopilots are prone to failure, subject to wildly varying conditions and they use a lot of juice. This popular accessory helps with operator fatigue but introduces a level of danger because of potential failures and operator complacency. Today’s pilots have alarms to warn the operator of failures and errors such as Off Course, Compass Failure, etc. As long as the operator sees the alarm and reacts to it appropriately, alarms are a big safety addition. Since autopilots are complex with many modes, training the user is imperative. There is no substitute for eyes through glass, an actual human watch-stander. Just because the autopilot steers for you, doesn’t mean you can take a nap!
