[TG] Autopilot efficiency

[gauvins]

Our boat is a heavy steel 13M cutter with hydraulic steering. It tracks well and has lots of inertia.

We have a raymarine SPX 30 course computer. I'd say that the pilot requires 4A on average. Which is a bit much as we wish to be 100% solar.

I understand that energy requirement is influenced by sea state. But maybe there are things that can be done to improve efficiency.

Two questions:

1. Is there a material difference between "smart" course computers and older generation (SPX 30). What order of magnitude?

2. How can we improve the efficiency of our SPX 30? For instance, even at low response setting (2), the pump (type 2 equivalent) is almost constantly active. Hand steered, we'd be much much less likely to change the rudder angle...

[Tom - Raymarine - Moderator]

Hello gauvins,

The newer pilots have a lot more information about the movement of the boat from one moment to the next, so make better decisions about when to drive and how hard, but how much of a difference this is going to make will be hard to quantify because I can't see when and why the pilot is driving at the moment.

In order to minimise the consumption by any pilot, some important things to check are:

• Is the fluxgate compass installed in a good location? As close as you can to the pitch and roll centre in order to minimise swinging in the gimball, and with as little magnetic influence from the boat itself (this is probably where the EV pilots will have the biggest influence on your boat: as a solid state, 3D magnetometer, they're not affected by boat movement and can be installed up high without the drop in compass accuracy from boat movement). Accurate heading data is absolutely critical to the pilot making the right course-keeping decisions.
• Is the compass well-calibrated, with a low deviation and accurate alignment?
• Is the rudder-reference aligned well, so it reads accurate and centred rudder angle?
• Is the drive well-matched to the boat weight and steering so that it has a suitable hard-over time (pilot's time to drive from lock-to-lock), within 8-12 seconds? Slow hard-over times mean that the pilot's constantly playing catch-up, whereas fast hard-over times mean that the pilot tends to over-shoot the desired position and oscillate.
• Sail trim
• Response level. I describe Response as the "don't care" factor in the pilot, where the lower the response, the larger the amount of difference between current heading and locked heading the pilot will allow before making a correction. Lower response will use less power.
• Appropriate pilot calibration for Rudder Gain, Counter Rudder etc.. I wouldn't change these yourself (though you can re-do the Autolearn process documented in the installation manual), but if these values aren't set appropriately for the boat and steering system then they can cause over-activity or poor steering performance, both of which will increase current consumption.
  

In short, the pilot needs the most-accurate possible sensor data, most-appropriate setup and calibration and best drive unit configuration in order to achieve the best efficiency.

If your pump is active almost all the time then I would guess that you have a problem with your sensor data, drive hard-over time/pump matching to steering cylinder or pilot calibration: that's not how the pilot should operate.

If in doubt, if you'd like to log 15 minutes' or so of NMEA2000/SeatalkNG data on a Lighthouse2 or earlier MFD (Axiom/LH3 can't yet do this) when you're using the pilot at sea, then post that back here, I will have a look at your pilot's activity and sensor data (heading versus COG, STW versus SOG, heading versus locked heading, rudder slew rate etc.) to see if I can see any obvious problems. If you let me know what model of (NMEA2000/STNG-connected) MFD you have, I can give you info on how to do this.

Regards,
Tom