Autopilot Problems

Our S1 Wheelpilot autopilot went on the fritz on the way up from Seattle to Port Hadlock, as noted here previously. The compass wouldn't track as the boat heading changed, an issue which manifested itself in gradual course changes off the designated heading, followed by sudden corrections in the wrong direction at odd times. It wasn't a big deal to hand steer and we had other things going on so I didn't mess with it on that leg of the trip, but I decided to look at it Friday morning as we were taking the boat over to Port Townsend for the haul-out. I had checked Raymarine's support website before we left, and found an advanced diagnostic for the sort of issues we were having.

I pulled out the multi-meter and checked the wire pairs for the impedance values specified in the diagnostic document, and sure enough, on one of them the measurement was four ohms above tolerance--probably a short. I re-swung the compass anyway, just to give it a shot, but when we tried the AP on the way to the boat yard it had the same issues.

I called Raymarine tech support while we were waiting for the lift. The guy listened to my explanation and the results, then told me to take the compass itself apart and measure the impedance values internally; if they were still high, the compass was probably bad and should be returned, otherwise it was a wiring problem.

So I took the compass head apart and measured the values on the wire pair again and got a reading not described anywhere or by anyone. I called Raymarine back, got another tech who told me the values were whacky and I should probably send the thing back in. This was not exactly music to my ears, since their turnaround time probably wouldn't get us the component back before we were due to leave. I had absolutely no excitement about hand-steering both ways to Alaska and back (Mandy didn't seem to mind, though... I think she hasn't had to do it enough).

But I thought about it a bit after I hung up with the tech, and the readings started to make sense--they were lower than the spec, but that made sense because I was measuring inside the compass, not across the wire. So for the heck of it, I went back and measured at the wire ends to the computer again... and the value now was dead-on what it was supposed to be.

All I had done was take apart the compass head, but apparently that did it. I'm still a little worried there is a short in it somewhere but it will be hard to tell unless it has problems again. After we got it back in the water and headed back to Hadlock, I tried it again. On the first try, it gave me the disturbing message "NO RUDDER" but I was pretty sure that I did have a rudder, since we were able to steer and all. I was worried the rudder sensor had somehow been damaged, but then realized it was more likely that I had disturbed the sensor wiring at the computer when I was messing around with the compass wires. I cycled the system off, checked the connections, turned it back on and got a good AUTO function when I tried it next. After that, it held a course all the way down the bay, no problem.