B&G autopilots

[GRIT]

Hi all.

I'm going to buy our autopilot and instruments soon, as it's boat-show season. I'd like to get some comments from anyone using B&G autopilots, preferably one of the newer ones, such as NAC-3 or H5000 (Hydra or Hercules).

It seems there are quite a few Outremers with the H5000 system, I'm guessing that it reacts more quickly than the NAC-3? Anyway, we've got a light, and presumably fast cat, so we'd like a quick autopilot to steer for us.

I only want to do this once, and not want to "upgrade" in a couple of years, so anyone with experience with both systems would have my undivided attention.

We've decided on B&G, so for those wanting to suggest another brand, it ain't going to happen.

Thanks for your time.
Paul.

[rom]

I have a heavy and presumably slow cat but ... I shall follow the same route within the next 2 years to replace my aging raymarine AP. Main reason is I miss an accurate wind vane mode downwind, second reason of course is I am getting afraid that my 11 years old AP shuts down while I am asleep ...

My recent experience with b&g has not been so good though. A friend of mine has it all on his brand new 2016 lagoon 450 and setting up the AP parameters was crazy. I was actually unable to help him and he has been through a lot of phone & mail exhcnages with both lagoon & b&g experts until someone finally managed to configure it properly.

It happened last winter in the carribean and the echoes I got is that many lagoon owners had problems with those new B&G systems. (before 2016 lagoon was installing raymarine). I really want to like those B&G systems though, so following the thread ...

[gamayun]

Please forgive me for opining about a multihull discussion, but from a monohull's perspective, I installed a new B&G computer with flux gate compass that hooked to an existing hydraulic ram drive (sorry, I always forget the brand specs of everything), and I absolutely love it. It replaced an older Raymarine system. The B&G is extremely responsive in big sea state and wind with water crashing over the deck and seriously reefed. Also responds better than when I'm helming. It's hard for me to understand how a multi could be so much different in AP response requirements. A friend and I installed it and there weren't any issues, despite all the horror stories I'd heard. Maybe just lucky?

[GRIT]

gamayun

I'm happy to hear your B&G autopilot is working well. Which model did you get, Triton2 or H5000? Autopilots are amazing crew.

Your experience is pertinent, thanks.

The main differences, as I see them, are: 1. Multihulls, and light, fast, monos, tend to be more responsive to the helm, so a slow pilot will be more noticeable. 2. The consequences, on a fast multi, of a steering error, slow response, or failure of the pilot, can be more dire.

More speed makes everything happen so much faster. Surfing down waves at near 20 knots is not the time to be using a slow thinking, or responding pilot. Even when not surfing, if your boat regularly sails in the mid teens, even if it were only during gusts, a broach will be unpleasant.

I feel that B&G has a lot of experience with fast boats, and have their algorithms well sorted out.

Rom. Autopilots can be time consuming to set up, no doubt. I'm usually tweaking mine for months, before I'm happy with the settings.

How has it been working for your friend since?

Cheers.
Paul.

[DeepCut]

I have not played with it, but my B&G (on Lagoon 42) does have a screen to adjust the responsiveness of the autopilot. I don't know how much compensation this does between the different models.

[rabbi]

@rom: In a 2016 Lagoon it must be previous generation electronics. The Triton 2 range was introduced sometime in 2016, IIRC it was late 2016.
Especially in terms of AP control the Triton 1 is already "old" these days, and I'll explain why.


Our first cat had a 1998 B&G AP with a 1Hz electronic compass. So all it knew was the current heading compared to what it was 1 second ago. The AP was pretty slow to react but still doing OK most of the time.

Next cat had a factory installed 2006 Raymarine AP with 10Hz electronic compass. Already much better and capable of handling the worst weather once it was configured correctly (it came with powerboat settings).
But it could only sense and react after significant changes in heading and could not anticipate any current or future course change based on movements.


Current generation APs "feel" the movement of the boat through a constant stream of sensor data about course, heading, speed, wind, rate of turn, pitch, heel, ... and can act before the boats actually changes course, just like a good helmsman does.
But... interpreting a ton of raw sensor data and converting this data stream into a stable course is a complicated thing. As this is new technology for small pleasure craft I expect every generation of APs to be better than the previous one. Better in terms of "converting raw sensor data into AP actuator control to achieve a stable course". Not necessarily better in terms of reliability and longevity, thats a different story.
No clue which manufacturer is ahead of the pack today, and this is likely to change with the software generations (not necessarily AP hardware). In a few years this learning process will be a thing of the past, and all APs should be more or less equal.



With regards to B&G:
In 2017 I replaced the failing Navman electronics on our 2008 Lagoon 380 with a set of B&G electronics, specifically Triton 2 instruments with N2K network sensors, Triton 2 AP control head and a NAC3 autopilot course computer, the Precision9 sensor and the N2K rudder feedback sensor.
I just bought the boxes from Simrad and did everything myself.

Once installed the commissioning is almost automatic. The instruments guide you through the shoreside commissioning for rudder end points, hard-over times, minimum & maximum power settings, etc. For seatrial commissioning the AP does a few pre-defined turns to auto-calibrate the rudder gain, counter rudder, etc.
Its all really all dead simple.


There is of course a ton of things you can configure for the AP, but in reality you just don't need to.
The steering has been reliable for us from the beginning in all wind or sea states. This setup has been excellent so far (one season and maybe 900nm).


The only complaint, and this is where I think further development will make a difference, is that it sometimes reacts too quick:
We are sailing with wind at 160 degrees, and one wave after the other arrives from the same direction. The human helmsman knows that the boat will turn a few degrees as the wave lifts the windward hull but will turn back as the wave passes. So he just does nothing, as both changes compensate each other for each wave that passes.
The AP on the other hand will compensate both of these small course changes. Not a problem, but simply unnessary activity.

I changed a few settings to reduce unnecessary activity, for example increasing damping for wind changes.


As said this is for the B&G Triton 2 with NAC3, feeding the standard L&S hydraulic drive that was factory installed on our Lagoon 380.

[GRIT]

Rabbi. I'm happy to hear from you. I was hoping I'd hear from someone, with direct experience, with the Triton2.

It sounds like you're pretty happy with it, and I'm glad to hear it can be fairly simple to set up. Though I'm pretty competent with electronics, it's always nice not to have to think too hard during commissioning.

Thanks for your input.

Paul.

[GRIT]

Deep Cut. Thanks for your comments.

There are significant differences between the H5000 units and the Triton2 units.
Many of the differences are race related, and deal with ancillary, and optional, sensors you can add to the H5000 units.
But I don't know if those differences will matter for my purposes. Nor do I know how large a difference there is in the algorithms that help predict which way, and how much, to turn the wheel 2 seconds from now. Perhaps the H5000 has the wave issue, that Rabbi spoke of, figured out. I don't know. But stay tuned, and perhaps someone who does know will chime in.


Thanks.
Paul.

[rabbi]

Originally Posted by svquintana

Perhaps the H5000 has the wave issue, that Rabbi spoke of, figured out.

No worries, this is just possible future optimization. The NAC3 is much better than for example a 2006 Raymarine, and that was already an AP that was considered good or excellent by most folks.


One thing: I just learned that one can still buy old stock Triton 1. These are much cheaper but don't by fooled ! They are not compatible with new generation APs like the NAC3.

I initially bought Triton 1 after asking the B&G service if they would work with a NAC3. During comissioning it turned out that they were wrong, and Triton 1 can't commission the NAC3.

Good for me: I received new Triton 2 in exchange for the Triton 1, but only because I had written evidence that it was a mistake by B&G pre-sales support.

[rom]

Hi Rabbi & thanks for the detailed explanation but you lost me here. I thought Triton was an instrument ? I mean not an AP. Right, my friend's 2016 L450S came with triton 1, but the setup was done with a zeus2. I do not see how triton or zeus are related to the intrinsics of the AP. I do not know which AP my friend has, he probably doesn"t know himself, I will try to find out.

Also you seem to say that 2016 b&g stuff is already outdated, sounds a bit hard for those who invested thousands in their electronic system !

I am glad you have good experience with b&g though. When in wind mode, do you change from AWA to TWA depending on you going up or downwind ?

[mikedefieslife]

Triton and H5000 are just instruments/displays

The autopilot computer is the NAC3 (the number differs depending on the power of the unit). The unit is the same for B&G and Simrad.

[GRIT]

The NAC-3 and the H5000 are different autopilots, the H5000 autopilot and CPU come to double the price of the NAC-3 autopilot.



An example of the H5000 system, which also includes H5000 branded instruments. For the pilot, you need the CPU, speed sensor, heading sensor, rudder sensor, wind sensor, and the Pilot computer at a minimum, to have a working system.
The Hydra CPU shown is the base H5000 CPU, there's also a Hercules, and a Performance CPU, which is nearly triple the price of the Hydra CPU.

At least that's my understanding until B&G returns my e-mail questioning that. With the NAC-3, you only need a controller, and the NAC-3 pilot computer, no separate CPU.

All very confusing, but almost definitely different systems.

Cheers.
Paul.

[seandepagnier]

Originally Posted by rom

Also you seem to say that 2016 b&g stuff is already outdated, sounds a bit hard for those who invested thousands in their electronic system !

Thousands? It cost only a few dollars for inertial sensors today.

I built my own autopilot, it can run on $5 raspberry zero. It is modular, components can be replaced individually.

Steered a trimaran at 15 knots. I have sailed hundreds of miles, it uses little power.

It uses all open standards, like signalk and nmea0183. Free software, you can fix it, and make it better, build your own!

I am writing machine learning for it, like self driving cars to make it very smart, but this is not working yet.

pypilot - open source marine autopilot

[rabbi]

Originally Posted by rom

Hi Rabbi & thanks for the detailed explanation but you lost me here. I thought Triton was an instrument ? I mean not an AP. Right, my friend's 2016 L450S came with triton 1, but the setup was done with a zeus2. I do not see how triton or zeus are related to the intrinsics of the AP. I do not know which AP my friend has, he probably doesn"t know himself, I will try to find out.

Also you seem to say that 2016 b&g stuff is already outdated, sounds a bit hard for those who invested thousands in their electronic system !

I am glad you have good experience with b&g though. When in wind mode, do you change from AWA to TWA depending on you going up or downwind ?

The Triton line refers to the display and the small AP control keyboard that is to be used with a separate display. But the choice of display is connected to the AP choice:

The 2016 Lagoon B&G pack contains Triton 1 displays. The Zeus MFD is an optional component for the Lagoon pack so the autopilot has to be commissioned using the Triton displays (in case there is no MFD). This has some impact on the AP choice since the Triton 1 can't be used for commissioning the NAC autopilots. So its unlikely the pack contains a NAC AP but a previous generation autopilot instead, like AC20 or AC40.
These "older" autopilots are not prepared to make use of the Precision9 sensors, only compass and rate of turn. At least that's what I read when I made my choice of electronics in early 2017.
Which doesn't make them bad autopilots, its just they are cut off from new developments.
If one needs these developemnts? I don't think so, my complains about the AP being too anally about course keeping is minuscle.



I agree its hard but the 2016 B&G pack is already outdated. Simply because Simrad have a new generation that is partly incompatible. Anyone with a set of Triton 1 (and no MFD) can't buy a compatible AP anymore.

The set of 2008 Navman electronics on my boat were outdated in 2009 when Simrad bought Navman and shutdown the Navman product line.
My investment in the 2017 B&G Triton 2 and NAC3 will also be outdated as soon as they announce the next incompatible generation. And they will, just to keep their cashflow happy.
Sometimes life sucks these days.


For the windvane mode, the AP automatically selects between AWA and TWA depending on upwind / downwind. I tried the fixed setting as well but found the automatic setting works best for us, especially in light winds.

[rabbi]

Originally Posted by boat_alexandra

Thousands? It cost only a few dollars for inertial sensors today.

I built my own autopilot, it can run on $5 raspberry zero. It is modular, components can be replaced individually.

Steered a trimaran at 15 knots. I have sailed hundreds of miles, it uses little power.

It uses all open standards, like signalk and nmea0183. Free software, you can fix it, and make it better, build your own!

I am writing machine learning for it, like self driving cars to make it very smart, but this is not working yet.

pypilot - open source marine autopilot

Yes a decent set of marine electronics costs thousands. Even if it contains parts for just a few hundret.

Your DIY approach is interesting for sure but not for everyone.