my ST6001+ steers 30° to starboard when I hit Auto

[Brann]

my ST6001+ is behaving very weirdly.

I think the problem appeared when I followed the compass calibration procedure (the 2 circles at 2 knots thing) because my readings were slightly off.

Here are the symptoms I have :
- when the autopilot is in standby mode, everything is fine. The compass readings are perfectly aligned with my GPS and my navigation compass.
- if I click on auto, the autopilot immediately steers 30° to starboard. ie if I head north and hit auto, the autopilot displays 0°, but heads to 30°
- if I click three times on -10, the autopilot displays 330° and goes north
- if I click back on standby, the autopilot immediately displays 0° (ie the current real heading)
- The autopilot regularly displays "off course" alerts. I wonder whether it's because he's comparing the course I asked him on the display (330° in my example) with the course actualy followed (0° in my example). My off course limit alert is set to 25°

This looks as a software bug, and I have no idea what to do next ; any hints?
Thanks a lot

[Jon Hacking]

The ST6001+ is just the control head. The "brain" is in another unit. For instance, we also had the 6001+, but our brain is a 400G (G for gyro equipped). Do you know what brain you have?

Also, there's another calibration procedure, where the autopilot system drives the boat in a series of sharp turns to set a number of parameters. I liken it to the autopilot getting a feel for the boat's handling characteristics. Did you do that calibration?

[electronicmarin]

the rudder is on the center on the display when you are going strait? the compass alignament procedure gave you less than 7 degrees compensation?
things that people seems not to undarstaind that betwwen the magnetic (filed with liquid) compass, the autopilot compass and cog (from the gps) you have 3 different things and that's normal, is basic navigation, the magnetic compass is without deviation, the autopilot one is in somme way compensating for that deviation (hence a different value) and the gps compass value cog (course over ground) is yet another value that's the true heading of the boat...
My advice: reset again the autopilot and with no modifications (just put the correct rudder type on st6001) see if the autopilot is keeping the heading when you press auto, then will see from there...

[Brann]

Here are a few precisions :
- the control unit version is 15 and the course computer version is ver 227
- my course computer is a raymarine smartpilot s3G ast
- my rudder is centered when I'm going straight


The autopilot worked flawlessly for thousands miles. Two weeks ago I connected it to my tridata to enable the wind mode. One week later (causality or correlation, I don't know), I noticed my compass was off gainst the GPS COG. This is not unusual in Brittany where I'm sailing because of very strong tidal currents, so I don't know exactly when it happened.

Anyway, I followed the calibration process (2 circles at 2 knots). I can't remember the compensation. The sharp turns configuration thing is for gyroscopic versions only, which I don't think I have. I think the pilot was behaving correctly then (but I'm not sure as I might have used my secondary auto pilot to go back to the harbour, I can't remember).

Next time I left the harbour, I had the very strange behavior I described.

To ElectronicMarin, I think I understand the difference between a true heading, a magnetic heading, and a course over ground. In my case I know my autopilot compass is reasonably accurate, as I compared it against other compasses, and even against a known alignement on ground. It is definitely not 30° off. I'll retry calibrating the raymarine to get the compensation it finds as I understand this is a useful debugging information.
When you suggest a reset, are you talking about a compass calibration reset or about a full factory reset ?

Thanks a lot

[Jon Hacking]

First of all, Brann (& ElectronicMarin) Welcome to Cruisers Forum!

The compass driving the autopilot is not magnetic (in that it's not a card suspended in liquid trying to align itself with the earths magnetic field). It's a fluxgate compass, which doesn't have moving parts (well, except that it's gimbaled inside its case). It should have a 4 or 5 wire connection to the brain. Were these wires disturbed recently?

The "2 circles at 2 knots" really just wants a very constant angular change, so it can work out intermediary angles. Therefore, it's not enough just to do "2 circles at 2 knots". You must do that step when there's no wind or current to disturb the boat's direction. Start the calibration cycle, then turn the helm very slightly (you may have to do it a few times to work out how much) & LOCK THE HELM at that angle. This should give you the constant angular change that the AP is looking for. The boat should complete 1 circle in 1-2 minutes, but keep going until the AP beeps that it's finished. (I suspect that this is where your real problem is, so you might want to re-run this calibration step - we have a very difficult time finding appropriate conditions for running this calibration step...)

I don't believe the second calibration is for gyro versions only. Also, if your brain is an "s3G", that sounds suspiciously like it DOES have a gyro. They're very simple now, just a single chip. Most smartphones have them these days. It was an expensive option when we bought ours in 2002, but now I suspect most AP brains have them. The next time you're out in relatively calm & open water, I'd run that second calibration step.

You bring up an interesting point about sailing in strong currents. If you set your AP to steer towards a waypoint, then it could easily point the boat 30 deg off your nominal straight-line course, to compensate for that current. But your COG should still be the rhumb-line to your waypoint. Note that it usually takes the AP a while to work this out, even though the RayMarine GPS calculates current if it has fluxgate & paddlewheel data available. But it sounds like you know this.

[Redline452]

For what it's worth, my ST6001+ (with 400 course computer and separate "Smart Heading Sensor" gyro module, so it thinks it's a 400G) occasionally skews off to one side (usually stbd) when first activated if it hasn't been used in a while. I never let it go far, always hit Standby when I see the helm turning rapidly.

A second activation of Auto works fine, so what I have found works is to press Auto-Standby-Auto in quick succession. I don't think the display shows the course offset yours does, so this may be something completely different.

[electronicmarin]

the s3G has a gyro, hence the G in the name, when i was talking about magnetic compass the fluid thing was to make myself clear that I'm talking about the traditional compass and i was not refering at the fluxgate...
short off topic: i've made a video on how to test the fluxgate on raymarine pilots:

https://www.youtube.com/watch?v=mC3Lri8uMog

the video is on a st4000 but the same values you have on all raymarine pilots with fluxgate... (not for the new EV serries).

[electronicmarin]

Originally Posted by Redline452

For what it's worth, my ST6001+ (with 400 course computer and separate "Smart Heading Sensor" gyro module, so it thinks it's a 400G) occasionally skews off to one side (usually stbd) when first activated if it hasn't been used in a while. I never let it go far, always hit Standby when I see the helm turning rapidly.

A second activation of Auto works fine, so what I have found works is to press Auto-Standby-Auto in quick succession. I don't think the display shows the course offset yours does, so this may be something completely different.

whenever you add a gyro to a smartpilot you have to do the seatrial calibration, that way the pilot knows how to handle himeself...

[electronicmarin]

Originally Posted by Jon Hacking

First of all, Brann (& ElectronicMarin) Welcome to Cruisers Forum!

The compass driving the autopilot is not magnetic (in that it's not a card suspended in liquid trying to align itself with the earths magnetic field). It's a fluxgate compass, which doesn't have moving parts (well, except that it's gimbaled inside its case). It should have a 4 or 5 wire connection to the brain. Were these wires disturbed recently?available. But it sounds like you know this.

when i said the compass the one with the liquid was for make myself clear that i was not talking about the fluxgate...

https://www.youtube.com/watch?v=mC3Lri8uMog

in this video I show how to test the fluxgate and diagnose problems fluxgate related on raymarine autopilots since is the same fluxgate for all of them (st4000, 5000, s1, s2, s3 spx ecc) not valid for the new EV serries...
Also i found on somme systems that the braid of the fluxgate wire was touching the metal part of the course computer, that made that disturbs where entering in the fluxgate circuit... the fluxgate braid (the negative) is separate and filtered ground from the ADC circuit so it must stay separate from the boat/autopilot negative...

[Redline452]

Originally Posted by electronicmarin

whenever you add a gyro to a smartpilot you have to do the seatrial calibration, that way the pilot knows how to handle himeself...

Yes, done. Also repeated from time to time, usually after spring launch.

[hamburking]

Did you recently move anything magnetic, or just big and metal near the "brain"? Think hard, any rearranging? Things like speakers, displays, wires with current in them, even the case for you ipod may have a magnet for the closure.

[LByers]

Any chance some sort of metal object was placed near the compass? I was taking a boat to the Marquesas when the autopilot started misbehaving. It turned that one of the crew had placed a toolbox in close proximity to the fluxgate. After relocating the toolbox things straightened out. Funny thing is, that same crewman stacked a bunch cans there later--same problem.

[hoolie]

But the OP said the displayed heading is correct so there can't be any magnetic disturbance near the flux-gate nor any calibration error. The autopilot just steers 30 degrees off. It does sound like a corruption of the software to me, certainly if the error is there for any heading.

[hamburking]

Originally Posted by hoolie

But the OP said the displayed heading is correct so there can't be any magnetic disturbance near the flux-gate nor any calibration error. The autopilot just steers 30 degrees off. It does sound like a corruption of the software to me, certainly if the error is there for any heading.

Its a modular system...the display may be getting a heading from a different source than the autohelm.

I had this trouble with the date/time on a similar system...couldn't get them all synced, drove me insane. There are "masters" and "slaves" in the system (autohelm's words, not mine, from the manual). You may need to map out your whole system to really understand what's going on.

[CampDavid]

Be sure nothing magnetic is close to tbe Flux gate compass. Tools or an anchor even moving the steel aux tiller can do it.

Did you can the wires to tge linear drive. If so reverse them.

Sent from my SGH-I337M using Cruisers Sailing Forum mobile app