Reminder Auto Pilot

[Alaskan Sea-Duction]

For those like me with AP. Lessons here?

https://gcaptain.com/fishing-vessel...139820873&mc_cid=43e7444404&mc_eid=5b70bcca05

 

[Gdavid]

Interesting situation, what I find most interesting is the AP failing to disengage.

The error of running an inlet while the helm is unattended, left on AP is straightforward and easy to find fault in the decision making but details within the story, that the captain returned to the helm and was unable to disengage the AP is easier to relate to. Having a switch dedicated to the AP that would kill all power to both the computer and drive is valuable but not always easy to configure with multi-station boats. I have experienced AP control pads with failing buttons, especially in exposed stations but that was one of the direction buttons, not the standby/auto button.

 

[Shrew]

Interesting situation, what I find most interesting is the AP failing to disengage.

The error of running an inlet while the helm is unattended, left on AP is straightforward and easy to find fault in the decision making but details within the story, that the captain returned to the helm and was unable to disengage the AP is easier to relate to. Having a switch dedicated to the AP that would kill all power to both the computer and drive is valuable but not always easy to configure with multi-station boats. I have experienced AP control pads with failing buttons, especially in exposed stations but that was one of the direction buttons, not the standby/auto button.

You might need to re-read the article.

According to the report, the vessel was transiting outbound [in the St. Mary's Entrance channel], the captain set the vessel’s autopilot to maintain the vessel’s heading out of the inlet. He answered a phone call and left the wheelhouse, but shortly after, the captain felt the vessel turn abruptly to port.

...

The investigation found that two days before the grounding, the captain was unable to disengage the autopilot and gain control of the helm as the vessel was proceeding into St. Augustine. The captain examined the autopilot system and found problems with the rudder angle indicator and rudder angle sensor at the rudder post and he took actions to correct the issues. That NTSB said that while the repairs worked initially, the vessel’s sharp turn to port indicated the system failed and the repairs were not effective.

 

[SteveK]

The sudden turn IMO would not be caused by the rudder position indicator or the inability to turn off the AP.
In a river channel I would set AP (and stand watch) to find over the same spot the boat would suddenly turn. On investigation I found there was a submerged electrical cable feeding power from one side to the other causing a magnetic change which disrupted the system. Once past there AP would continue normally.

 

[Portage_Bay]

The sudden turn IMO would not be caused by the rudder position indicator or the inability to turn off the AP.
In a river channel I would set AP (and stand watch) to find over the same spot the boat would suddenly turn. On investigation I found there was a submerged electrical cable feeding power from one side to the other causing a magnetic change which disrupted the system. Once past there AP would continue normally.

I have had the problem of the boat going hard right due to a bad rudder position feed back unit. When the AP corrected to the left all was good. When it corrrected to the right the rudder went hard right and stayed there.

The article used the phrase disengage not turn off. If, I realize that is a huge if, the article's use of the word disengage is accurate that implies a different kind of system than most of us use where killing the power restores manual steering. Possibly a continuous running system where a problem with the solenoid shuttle valve leaves the system in a bad state such as hard over. Killing the power or turning the AP off will not restore manual steering becaue the valve is stuck in the rudder turn position. If there is no manual steering then the fall back to AP is a jog stick, that won't work either when the valve fails. The boat looks to be an older heavy duty boat. It could have a system so old that a clutch is used like in the old Wood Freeman systems. Clutch disengage failures are extremely rare but I suppose it could happen. Of course this is all speculation, something we here on TF are very good at.

We'll never really know because the boat broke up and sank.

His primary error was leaving the wheel house unattended. Had he stayed there he could have had time to deal with the situation.

 

[SteveK]

I have had the problem of the boat going hard right due to a bad rudder position feed back unit. When the AP corrected to the left all was good. When it corrected to the right the rudder went hard right and stayed there.

I have not had a rudder position indicator.
But I cannot understand why an AP would turn the boat if it is on course because a rudder indicator tells it is not centered. Sometimes a rudder is off center to keep a straight line.
Look forward to understanding. Maybe I will not get one if it has this issue.

 

[psneeld]

Any sensor can go bad, some induce full correction if the sensor indicates in error until removed.

Hook one up backwards and see what happens.

 

[SteveK]

Any sensor can go bad, some induce full correction if the sensor indicates in error until removed.

Hook one up backwards and see what happens.

Agreed, but my study of the rudder sensor, and the reason I have not had one is that it is only telling where the rudder is so the AP knows what point it is full over, so not to over steer. Never saw anywhere that it controls steering. So I thought it can show hard over one way all day until AP wants to turn that way, it would not.

 

[psneeld]

Agreed, but my study of the rudder sensor, and the reason I have not had one is that it is only telling where the rudder is so the AP knows what point it is full over, so not to over steer. Never saw anywhere that it controls steering. So I thought it can show hard over one way all day until AP wants to turn that way, it would not.

Depends on AP software. What it senses and how the AP uses it.

https://marinehowto.com/installing-...rse.?msclkid=41f913a2d15711ec8a622aa2d8fffcae


A rudder position sensor is really just an autopilot system transducer. The rudder position sensor (RPS) tells the autopilots course computer where the rudder is at all times. The RPS helps the AP’s course computer in holding an accurate course. Over the last ten or so years AP manufacturers have tried very hard to program out the RPS by adding features such as rate gyros, prediction & behavior learning software etc. etc.. Unfortunately with all this technology nothing improves AP performance like installing a rudder position sensor.

On low end price point autopilots, such as Raymarine’s wheel pilot systems, the RPS, which used to be standard equipment, is now optional. In my experience the RPS is not “optional” if you desire the best performance from your AP. For coastal cruising you can sneak by without one but your AP performance will always be better with one.