Wacky Auto Pilot

[Larry Robertie]

Hi folks,

I was out on the boat today trying to get some solo time in.  The good news is that she sails great, 20 knot winds, full sails and all. 

The this was though that throughout the Spring I had noticed the autopilot as being off by some 30-35 degrees.  Not having had an autopilot in the past I just left that to the last to experiment with so...

Today I'm sailing out of Salem harbor with the motor idling just in case.  The auto pilot is off by 30-35.  About half way out of the harbor I shut down the engine and a little later I notice the auto pilot is pretty much within a few degrees of the compass.  I think, well maybe it just had to warm up. 

Sailing with the auto pilot worked well.  When engaged it kept the course.  It even tacked properly, until it came on course at which time it had trouble holding.  But I chalked that up to the 20nm winds and the fact that I had all the rags fully out.

Comming back in I decided to hall down the sails and motor.  Low and behold the auto pilot went off by the 30-35 degrees.  And, something I had slightly noticed before, the integrated chartplotter display is also off from th e"heads up" orientation - by 30-35 degrees.  I turned the auto pilot off and instantly the chatplotter displayed correctly.

Checked the fluxgate compass and there was nothing near it.

About three weeks ago I was out following the manual and doing the calibration thing on the autopilot buy got a message back that the deviation was 26 degrees.

Anyone had a problem like this?

[tommyt]

Larry,

In assume that this is a ST 4000. I have one and hate it, but that does not mean that every 3rd one does not work.

In calibrating it should not be over 12 degrees so you need to do it again. Mine also gets devitation over time and I have to go back and correct to compass setting or I wander all over the place. I have a thing about drinking and driving ( note I DID NOT say drinking), but you would swear that when the autopilot is out of whack I was stoned. Looking back at my wake I swear it looks like I am making turns of 5-10 degrees as I move along.

[Ken Heyman]

Larry,

I concur. I have had to go through the calibration procedure more than once to get the deviation within acceptable limits (i.e. spinning the boat & following the manual's instructions). Flat seas do help. Also don't forget to "save" the settings or the unit will default to the previous settings and the entire process will need to be repeated.

Good luck,
Ken

[Jack Hutteball]

I have never had the luxury of having an auto pilot, but on our 27 I noticed that when I ran the diesel engine the compass would swing approximately 15 degrees off course.  I assumed it had to do with wiring running somewhere close to the compass.  I never worried about it, just compensated for it when I needed to.  I soppose this could also be the cause of your fluxgate compass error?
Jack

[dbpaul]

I just reinstalled my, 2002 year, autopilot in which Raymarine repaired by installing a new PC board.
Went around in circles and came up with 8 degrees off set. I made the adjustments per the manual.
So I'm sailing along with the autopilot engaged every thing looks OK autopilot makes the good noises
making 1-5 degree corrections then less then 5 minutes later it went off 30-35 degrees and would not make a correction back to the original course.
I made a clutch adjustment when I got back to the slip and found nothing unusual.
Will be out tomorrow in Guemes channel which has a lot of eddies when tide comes in for a longer series of tests.
paulj

[Larry Robertie]

Thanks for the replies so far.  Nice to know I'm mot allone with this problem.  Looks like I may be doing the circular calibration thing some more, but I'm still perplexed about the variance between the engine on and off.  Of course I do not have a great many observations and could be drawing the wrong conclusion, but there did seem to be a strong correlation.

[sedelange]

The offset you are getting sounds interference from the alternator or some other engine circuit.  In airplanes we could just isolate the alternator circuit.  With the Catalina I think the only option would be to turn off the keyswitch momentarily to see if the error changes.  If it does, you might have wires to the engine paralleling the wires to the autopilot.  Moving the wires where the cross at 90 degrees could help.  Installing a filter on the alternator is another possiblity.

[Larry Robertie]

Thanks Steve.  Sounds like the old days when we needed to put shielded spark plugs in car engines or they'd reek havoc with the radio!

[Steve Sayian]

Larry,

I had the same problem on my '99 until I talked to another guy in the Club w/the same problem after he installed a new 4000 on his Tarten.
The problem is that your configuration setup file is set for mechanical instead of hydraulic steering.
I know it sounds nuts since you have 'mechanical' linkage, but the issue is the number of turns the wheel has to physically make from stop-to-stop.  This is what Raymarine told my friend.  My system did the big wandering thing and when I made the change, she was steady-on to within a degree.

It's real simple to change.  When I get home tonight, I'll look up the procedure and send it to you.  You do the whole thing right at the instrument head.  Takes about 10 mins.

Steve

[rirvine]

Also see the discussion at:   http://c34.org/bbs/index.php?topic=3043.0

Ray

PS this has some links to other posts, including the Catalina Cocktail setup for these units [Stu]

[Ken Juul]

With the engine running,initially the biggest load goes to the batteries until they are back to fully charged.  Is your flux gate or it's cable in close proximity to any battery cables?

[Larry Robertie]

With the engine running,initially the biggest load goes to the batteries until they are back to fully charged.  Is your flux gate or it's cable in close proximity to any battery cables?

Good point, and yes it is.  The FG is bolted to the aft part of the bilge - just to port of the batteries!  Looks like a move may be in the near future.

[sdaly66]

I had a similar problem on Friday.  Suddenly went offcourse and wouldn't correct.

Could this be a power interruption issue?  Because she generally work wonderfully.

I've used her to tack (I only have the compass option - not apparent wind), but it worked like a charm.

Just this one-time glitch for which I cannot account.

[Ken Juul]

One other possible cause of unplanned course changes is if the VHF antenna line runs too close to the Fluxgate.  When transmitting on the mast mounted VHF antenna I get an immediate 40 degree port turn.

[sdaly66]

Ken,

I thought that was a FEATURE, not a problem !   LOL