Larry Terrace on Pacific Psalm

Larry Terrace guides Pacific Psalm his beloved 1998 36 foot Catalina MKII away from the Royal Vancouver Yacht Club (RVYC). We were about to re-calibrate and sea-trial his dated ST4000+ Autohelm autopilot that hadn’t worked properly for nearly two years. It had developed this habit of veering wildly, and neither of us knew how to fix it…

Royal Vancouver Yacht Club

Back in May my friend Warren Terrace asked me to come down to the RVYC and see if I could help his Dad figure out the autopilot problem. Larry had asked his mechanic and the techies at a local marine electronics store but neither were able to come up with a suggestion that helped. I wasn’t confident and was ill-prepared to investigate older electronics I knew nothing about. The tentative plan was to get more information and then perhaps research the problem or ask knowledgeable readers.

 Pacific Psalm Electronics
Pacific Psalm Electronics

Catalina Electronics

Larry uses his Catalina 36 for local cruising and it’s a popular model with more than 3,000 built from 1982-2006. Here is a picture of some of his electronics topside. When I stepped onboard I knew I was looking at gear that had been on the boat a long time. The Autohelm/Raymarine displays and ST4000+ autopilot control unit in the main cabin were likely installed within a few years from when the boat was built. Larry had bought the boat used 8 years ago and the electronics were already installed. He uses the Raytheon RL70RC as a chartplotter but hasn’t learned how to use the connected radar. I noticed his C-MAP chart card was seven years old. Time for an update!

That empty holder in the lower left of the photo was for an ST600R wired autopilot controller that Larry thought might be causing the problem. He had upgraded it to a wireless SmartController (A18105) in the hopes of fixing his veering steering problem but it hadn’t. Today the SmartController batteries were dead.

ST4000 Autohelm on electrical panel

The electrical panel was partially open with a wire leading out of it to a Raymarine fluxgate compass nearby. Larry mentioned he thought the old fluxgate compass mounted permanently in the head wasn’t working. He had gone out and purchased a new one which he had placed horizontally on the countertop in a temporarily installed state. I didn’t think this was what Raymarine had in mind when they put a vertical arrow on the back of the mounting plate. I suggested we re-connect the original one because it was vertically mounted on a bulkhead.


I’ve always found it helpful to try to remove as many variables as possible before tackling a problem. A good starting point is to try to revert things back to exactly the way they were when the problem started. I opened the electrical panel further in an attempt to find the termination of the original wires.

electronics panel wiring

It was a bit scary behind the panel but I did recognize these three Seatalk connectors. After some fumbling around I located the 5 old fluxgate wires and substituted them on the back of the ST4000+. I then removed a Seatalk connector which led back to the wireless base unit and re-connected the old controller.

ST600R autopilot controller
ST600R autopilot controller

The wired controller display was in rough shape but the display powered on and showed a wild GPStimate of course over ground while we were stationary at the dock. Magnetic bearing readings from the fluxgate compass were not accurate in comparison to the compass on the binnacle, so I started wondering if that was contributing to the problem. I thought the next logical step was to check the settings menu on the ST4000+ (software version 14) against instructions in the original manual which Larry had available. Some of the settings looked logical but others like rudder gain — set to 9 — seemed pretty aggressive. I changed a bunch of things and set gain to 5 as suggested in the manual. We headed out for testing…

Sea Trial

The first step was to perform the automatic deviation correction sequence in the setup menu which involves doing a three minute 360 degree turn at least twice. Once we had this complete (6 degrees calculated) I noticed that the electronic magnetic heading differed 25 degrees from the binnacle compass. I corrected it in the setup menu to better match up.

Map showing autohelm testing area near Jericho
Testing in English Bay – Vancouver

At slow speeds we activated the autopilot on a fixed heading towards downtown Vancouver using office towers as a visual reference. (approximately shown in the above Google map). Despite what seemed like a lot of unnecessary corrections by the motor drive, it generally kept us on the same course. After speeding up to 6 knots things were fine with manual steering but when the autopilot was set to hold a specific bearing the boat started to zig zag aggressively which was pretty disconcerting.

On the handheld I looked at the gain and it was at 9 for some reason. I’m still not clear why it didn’t match to the ST4000+ display but perhaps the remote’s internal setting overrides it? I set it back to five and the autopilot seemed tamed, no sudden turns and it was making satisfactory course changes using the 1 and 10 degree buttons. It still seemed to be working awfully hard but Larry was satisfied. Back to the dock.

ST4000 autohelm wheeldrive actuator diagram of belt change
wheeldrive schematic


After my visit I did some online research and discovered this fantastic entry called taming the ST4000 by Jack Brennan. After reading it I passed it on to Larry and suggested he investigate the belt (#5 on diagram) inside the Autohelm wheel drive unit. Perhaps it was belt slippage causing the autopilot to work far too hard for my liking. I also sent Larry a copy of Raytheon’s ST4000+ autopilots service manual for his reference.  Sadly, over the next few cruises the Autohelm returned back to its aggressive ways. The autopilot was not yet tamed!

Fast forward to the end of the summer and a new drive belt was finally ordered and installed. I’m happy to say that this completely solved the problem!  All along it wasn’t the electronics!

Besides updating his charts I’ve suggested Larry upgrade his multifunction display, perhaps a 9 inch Raymarine aSeries touch screen?

Pacific Psalm at RVYC