SailNet Community - View Single Post - Fine-Tuning the Autopilot—Part One
View Single Post
post #1 of Old 09-03-2003 Thread Starter
Dan Neri
Contributing Authors
Join Date: Jan 2000
Posts: 5
Thanks: 0
Thanked 0 Times in 0 Posts
Rep Power: 0
Fine-Tuning the Autopilot—Part One

The author's Aerodyne 38 has roughly 10,000 miles under its hull, many of them logged while Auto was minding the helm.
I have actually heard somewhere that autopilots have the propensity to learn, gradually zeroing in on the correct amount of rudder movement. That doesn't sound right to me and I have not seen any evidence of it. But autopilots do have the ability to teach. If you pay attention to the differences in the boat's sail trim when the pilot is behaving well and when it is behaving badly, you might start to learn what sail shape combinations make your boat steer straight ahead, and what shape combinations result in the snake wake.

Unless you have ordered a very specialized, custom pilot for your boat, the manual that came with the one you bought will include several pages describing the standard adjustments and calibrations. The pilot that we installed on our Aerodyne 38 is an Autohelm 6500+ with a below-deck, electric, linear-drive ram. While this is one of the more popular models for boats in the 40 to 50-foot range, I am pretty sure that I am the only person who has ever read the entire manual. Over two years and about 10,000 miles of sailing I tried all the different settings and eventually learned that the factory defaults were just fine. In fact, if I had just tossed the manual in the storage bin along with the instructions for the stove and the head and the fuel tank sending unit, I would have been better off. The plasticized card with cartoons that Autohelm supplies for the American market turns out to be the only source of information about adjusting the pilot that I have needed so far.

What I have learned is that no matter how well dialed-in the pilot's brain is, if the sails are not set up correctly for the wind conditions, the boat won't sail in a straight line. With that in mind, the easiest way to set up the boat for the pilot is to put the best helmsman in the crew on the wheel, trim the sails until he says the boat feels right, and push the AUTO button. If you have set the pilot to the factory recommendations for your type of vessel, your boat will get to where your best helmsman has just aimed it and no one will get hurt along the way. If you were lucky to get it just right, the boat will sail in a very straight line. If the sails are out of balance or you are sailing in a reasonable amount of wind and waves, the pilot will cause the boat to proceed by leaving S-curves in its wake. That's when you know it's time for some fine-tuning.

Large diameter wheels, like the one on the author's boat shown here, are often a challenge for auto pilots due to the inertia they can develop in washy conditions.
To see what happens when a boat is out of balance, trim your sails for a close reach, engage the autopilot, and then dial it down to a broad reach. The pilot will turn the boat to the prescribed compass direction and then bring the rudder to centerline. With the rudder on center, the overtrimmed sails will turn the boat into the wind. Then the pilot will steer back to course and the cycle will start over again. Assuming that the boat is sailing in moderate conditions, there are two courses of action to take to remedy the S-curves; turn up the power to the pilot or balance the sail trim.

Power Settings    Most pilots have a range of power settings available. On my Autohelm pilot I can choose between three "response levels." In response Level 1, the pilot uses very little battery power, but there is a longer delay between each adjustment to the boat's course. Waves, excessive heel, or badly trimmed sails cause the boat to turn. The longer it is between rudder responses to off-course steering, the farther the boat is allowed to turn and the bigger the S-curve will be upon correction. So the idea is that if you are operating in flat water and very light air, you can conserve power by using the lowest response level.

Let's think about that for a second. On a cruising sailboat, in light air and flat water you will most likely be motoring. When motoring the alternator is pumping up the batteries at a far greater rate than the pilot is draining them, so you might as well step up to the next response level, and steer a straighter course and save yourself time and distance.

In practice, we pretty much always turn the response level immediately to Level 2. Even with the sails perfectly trimmed, on a light displacement boat like the Aerodyne 38, there will be some amount of S-curving in the course. And the bigger the S-curves, the farther the boat has to travel to get to the other end of a straight line.

Setting the response level is important whether you've got a linear drive unit or a wheel pilot like that shown here.
On the Autohelm, response Level 3 engages the optional Gyro sensor. The Gyro picks up any changes in the rate of acceleration of the boat around the compass. In other words, if the boat is steering at a steady rate to port, the normal pilot will be employed and will respond by repositioning the rudder to counter the turn. But if the gyro senses that the rate of the turn is increasing, more angle will be applied to the rudder adjustment. The best gyro units sense change in the boat's pitch, roll, and yaw as well. Your brain does that too when you are steering the boat and looking at the attitude of the bow relative to the horizon. Your brain also does that when you are inside the boat reading the autopilot manual at the nav table. That's when you get into trouble because your eyes are focused on the book and they are telling the brain that everything is cool. Meanwhile the brain's gyro is pretty sure the world is changing angle in three dimensions. After a few minutes of these conflicting signals, the stomach gets confused and tries to send its acids outside for a look around.

Gyro Issues    When the gyro is employed the pilot's ram is in constant motion. That puts a strain on the ship's power for two reasons. Every command from the pilot brain to the ram requires a pulse of current so the more commands per minute, the more power is drained from the batteries. Also, the commands for ram movement primarily involve changes in direction of travel rather than incremental movements in the same direction. A movement of the rudder blade to port involves not only the rudder and shaft, but also the entire steering quadrant, linkage system, and wheel or tiller. The next movement to starboard requires enough power to reverse the flow on the rudder blade in addition to the power required to overcome the considerable inertia of a spinning wheel.

"Watching the gyro slam the wheel back and forth in surfing conditions we often wonder if the pedestal is going to simply rip out of the floor."
A heavy displacement boat doesn't get chucked around in waves as violently as a lighter boat so there are less changes in direction required for the pilot to keep it going straight. At the same time, heavier boats typically carry larger, heavier battery banks so power consumption is less of a concern. Our boat is the worst-case scenario for gyro steering—a light, beamy boat with high-aspect foils and a big sail plan. Add to that ensemble a large diameter steering wheel and it gets a little unnerving. Watching the gyro slam the wheel back and forth in surfing conditions we often wonder if the pedestal is going to simply rip out of the floor. The obvious solution to the problem of wheel inertia is to disconnect the wheel from the steering linkage when the pilot is engaged. That's a good solution until the pilot inevitably cuts out, or gets inadvertently switched off. Our solution is to only use the gyro setting in rather extreme conditions for the short period required to match the sail plan to the new wind strength.

Steering Input    With a NEMA interface and an apparent wind angle sensor, the autopilot will offer three choices of steering input; compass, apparent-wind angle, or course over ground. Apparent-wind angle is the best choice for sailing hard on the wind in open water. On reaching angles with a cross sea the boat tends to hunt a little more in wind mode as the waves push the bow around, so the compass setting is a better choice.

If your sails aren't properly trimmed, and there's any significant amount of wind or waves, you'll likely need to fine-tune the settings on your autopilot.

But sailing downwind in moderate conditions really presents a toss up when it comes to selecting the steering input. If the boat is surging, the pilot seems to do well in wind-angle mode. As the boat accelerates and the apparent wind moves forward, the pilot will sense the change and steer away from the wind. Problems start to occur when the conditions change from surging to surfing. Aboard a light boat in big waves the pilot is tricked into bearing off too far as the boat accelerates down a wave. When the boat approaches the trough and slows down, the wind swings abruptly aft and the pilot tends to overcompensate with a sharp turn toward the wind, just before the sails load up and spin the boat that way as well. An extreme example of this can be found aboard an Open 60 single-handed racing boat in southern ocean waves. These boats are capable of speeds in the 20 to 25-knot range, under autopilot, when they take off down the face of 100-meter-long waves. If they are steering using apparent wind input the boat can bear off far enough that when it hits the trough and decelerates, it ends up on the wrong jibe. If the boat is surfing, it is usually better to use the compass mode. If you've got a spinnaker up at the time, it may collapse more frequently in this mode, but when it refills it will not be with such violent results. To my way of thinking, if you are sailing with both a kite and an autopilot in ocean surfing conditions, it might be time to review of your priorities.

The course-over-ground setting, or "GPS Mode," makes sense when sailing or powering in a steady sea state or in a coastal area with changing current. The problem with using a GPS for steering is that the COG calculations of the GPS are imprecise based on the differential error of the signals. For navigation purposes, GPS users typically set their units to a high dampening level so that the error is more or less averaged out of the reporting. For pilot steering, the dampened GPS input is fine in a smooth water situation, but will not make course changes with enough frequency to correct for wave action.

Dan Neri is offline  
For the best viewing experience please update your browser to Google Chrome