Raymarine ST2000+ auto pilot - was erratic - now just won't hold course

[Modulation]

It's about 3 years old and worked fine until the end of last year when it would suddenly - and violently - steer right off course. I talked to somebody at Raymarine who advised recalibration and that worked. Turned out that the setting was for southern hemisphere ( tho' why it had defaulted to that we don't know)
Anyway it's started going off course again. We're getting compass readings at the head so the fluxgate must be ok? I've tried recalibrating again but it's not sorting the problem this time. I've talked to a Raymarine repair agent who feels that the solution might be a new pcb (160 + vat + labour) which could get close to the cost of a new unit - and it still might not be the answer.
Any one had similar experience and suggestions on what to look for before replacing it?

Thanks all

 

[chewi]

could be a weak power line if i hasn't got the grunt to keep course, but if it has plenty of grunt but applies it badly that's prob. not it.

Has it got wet? I fixed one for a friend, I found some salt crystals on the PCB, I dry brushed it & tried again, but not very expectant of success, yet it worked.

 

[macd]

A possibility rather than a probability, but I had one doing the same thing, and it turned out to be section of broken rubber O-ring which had come adrift and was sitting neatly in the bulge below the compass gimble, intermittently preventing it from moving freely. The other odd thing was, I never traced the source of the O-ring.

 

[DaveS]

A possibility rather than a probability, but I had one doing the same thing, and it turned out to be section of broken rubber O-ring which had come adrift and was sitting neatly in the bulge below the compass gimble, intermittently preventing it from moving freely. The other odd thing was, I never traced the source of the O-ring.

Snap! I had exactly this experience a few years ago. Also no firm idea where the broken O-ring came from, but guessed from its size that it was on the inside end of the ram, maybe a bump stop? The plastic bearing housing there failed (spiral fracture) later, and I added a carefully shaped aluminium plate bolted on to reinforce it. Currently working well.