Raymarine ST4000+ Autopilot: Installation

When I acquired Sunny Spells, she had a Raymarine ST4000+ wheel pilot installed. The autopilot was only about 18 months old, as evidenced by the invoice which was still on the boat. The autopilot was an important selling point when I bought the boat, because I knew that I would be sailing short-handed or solo a lot of the time.

Unfortunately, I had to resign myself to the fact that the Autopilot was a fair-weather sailor only! It would helm fine to windward on a nice day (i.e. winds in the 10 to 15 knot range, flat seas…), exactly the conditions in which I could trim her to sail hands-off anyway! So what good is that then, you may well ask?

Well, in short, none at all…

To my frustration, the autopilot would snake all over the place when going downwind, often disconnecting because it couldn’t keep the heading within 40 degrees. Start the motor, and the heading on the fluxgate compass would change by up to 30 degrees, and changing the throttle setting would have similar effects.

I assumed that this was the sort of performance available from an entry-level autopilot…

During the refit that followed our dismasting, I decided to fix a couple of things that might affect the performance of the autopilot:

  • the rudder angle sensor was never fitted, but it was on the boat, complete with all fittings;
  • the fluxgate compass was originally installed about 12 inches from the alternator/engine… Need I say more?
  • the power supply to the Autopilot was an unprotected connection direct to the house bus, used un-tinned wire and was connected with those screw terminals used in home wiring. Corrosion everywhere…

The re-wiring of the electrical system took care of the power supply side of things.

The fluxgate compass was moved to one of the intermediate bulkheads, under a bunk and next to the flexible water tank. This location was close to the keel (i.e. a stable location with the least movement but accurate representation of the boat’s heading) and far away from potential interference (engine, alternator, vhf, loudspeakers…).

The rudder angle sensor was installed, but I could not get it to work. A quick check with the multimeter indicated that it was dead as the proverbial.

I assumed the sensor had to be a simple potentiometer, but none of the wire pairs would measure any resistance on the multimeter. I took a deep breath and opened the unit: two wires broken off at the solder connections to the potentiometer! Nothing a bit of solder wouldn’t fix…

After re-connection, the autopilot still wouldn’t show rudder angle!

Out with the user manual…

After a lot of trial and error, mostly error(!), in desperation, I told the autopilot that it was now a “tiller-pilot”, not a “wheel-pilot” (the tiller-pilot can’t accept rudder angle sensor input…). Lo and behold, as soon as I saved that setting, the rudder angle appeared on the display! I restored the setting to “wheel-pilot” and the rudder angle was still displayed!

Now for sea trials…

The sea trials exceeded my wildest expectations! The heading info from the fluxgate compass is now stable (no interference when running the engine) and, most importantly, the autopilot now does a great job of steering the boat under all conditions: upwind, downwind, reaching, running. I’m also impressed with how it auto-trims so that it keeps course better and better the longer it is engaged. Moral of the story:

  • autopilots need a high capacity, clean power supply;
  • choose the installation location for the fluxgate compass very carefully – this is the most important input the autopilot has; and
  • install the rudder angle sensor, even if it is a pig of a job (not easy to get to the rudder post…).

[ad]