Jeff,

  That pypilot is amazing. 

  I can't think of a system that provides more bang-for-the-buck.  That would make it easy to afford a six-pack of tillerpilot motors.  Then two pypilots, one for each tiller.  With some reprogramming they each could go dormant and switch from one to the other after going backwards for X seconds.

 That said, I've spent the past few days pondering this and looking at the setup, and am considering using the pypilot as part of a more expensive system.

  Basically: 

    - A single traditional nmea2000 or seaTalk (pardon my Raymarine reference) autopilot.
    - One or two nmea2000 chartplotters.
    - Two active tillerpilot motors.
    - One TackTick wireless wind/speed/depth system with a mast rotation sensor.
    - Two each of the following:
        - Autopilot 9-axis compass/sensors
        - TackTick fluxgate compass
        - TackTick speed/depth sensor
        - Nav lights

    - And one physical switch to toggle wiring connections between the two sets of sensors, and also the two tillerpilot motors.

  Then...

    - A pypilot with some reprogramming to flip that physical switch

  The pypilot could easily be reprogrammed to turn a controller motor one way or the other after it senses a new direction for X seconds, and that motor could flip the switch.  The motor could be anything -- an actuator, a winshield wiper motor, whatever.

  As a plus, you could disconnect the actuator from the physical switch and flip it manually if the pypilot stopped working.

  The wind, speed, and autopilot sensors would always think the boat is heading "forward", and the rest of the electronics would process their signals accordingly.  You'd get true or magnetic heading on the tacktick instruments (adjusted for declination based upon location determined by the gps chartplotter), true and apparent wind speed, and autopilots and a chartplotter that agree on what forward means.

  You'd also get the option to play with which rudder you want to steer with -- just flip the wires if you end up liking aft more than forward.

  Another bonus is that there's some redundancy built into the system, but without having to duplicate the expensive parts.

  (Then perhaps some extras could remain in sealed containers, with silica gel, inside a Faraday cage, as spares).

---

  As for Ryan's reasoned preferences, it would also work with two beefier linear tiller actuators, either mechanical or hydraulic, mounted inside the leeward hull with just the ram protruding. And/or a wired wind sensor (I wouldn't put one on a sail mast, but there's no harm in a wire running to an 8' non-rotating instrument mast.

  One neat thing about the HP is that an instrument mast on the windward hull would actually get clean wind almost all the time.

        - Mike

 

(I wrote this on a plane a few days ago but looks like it didn't go through, sorry if this ends up being a double post)

I've been giving the overall subject a bit of thought myself, though mostly thinking of tiller pilots and wind vanes.  HP's are still abstract to me, but I've read several notions of setting the rear rudder for balance and steering with the front. 

With that in mind, a tiller pilot on each rudder seems feasible, even if only the front one is active for either direction.  Of course tiller pilots arent known to be as reliable and sturdy as a good hydraulic system, but hydraulics seem a bit overkill for a 40ish footer I'd be interested in.  Dual windvanes would be both expensive and busy.  Not to mention clumsy if you had to switch on each shunt.  So its good to hear the 18er works fine steering either rudder.

Since I will be building my HP, I quickly fell down the rabbit hole of self built wind vane designs which could be feasible for either or both rudders.  Looking at simple self built designs made me embarrassed to see what hydrovane charges. 

In the true spirit of DIY, the pypilot project comes to mind.  Its all open sourced hardware and software for autopilot via openplotter on a raspberry pi, so theoretically someone (not me anytime soon!) could sort out the software side of bidirectional autopiloting.  The logic at least doesnt seem too difficult to put into human words..  Having a little experience with Raspberry Pis and home automation I can actually imagine several proa applications, such as automated nav light switching.

Whatever I end up with it will be DIY and/or used and steeply discounted.  With the economy built into HP's you could spend as much on all the modern gizmos as you do on structure.  Yachting is predominantly a rich man's game, but the underlying tech need not be.

On Tue, Dec 4, 2018, 1:02 AM Rick Willoughby rickwill@bigpond.net.au [harryproa] <harryproa@yahoogroups.com..au wrote:

 

The 18m proa was fitted with hydraulic steering two years ago.  That enables fitting an autopilot with hydraulic pump to the system.  The system operates one rudder and that works fine for steering in either direction.  We have only played around with the autopilot twice.  The incorporated GPS and radar have been more useful to date than the autopilot.  We are yet to resolve how the autopilot can be set to work in reverse without recalibrating each shunt or have a reversing switch on the pump motor.  There may be a simple software switch for the reversing function but that has not yet been found if it exists.  Generally manufacturers of autopilots have not contemplated their use on a proa.  

Setting the pump up was a simple exercise with the existing hydraulic steering.  The pump has in-built check valves so it does not need to be isolated to steer by hand but the helm needs to be isolated so the rudder turns in preference to the helm spinning.  The helm could be locked in place but it is simpler to use an isolating valve.  

The hydraulic steering was quite simple to set up and not particularly expensive using the same orbital pump/motor for the rudder and helm.  The steering design would be a bit different if done again but the system has proven reliable.  

Rick

On 4 Dec 2018, at 4:18 pm, ryanonthebeach@gmail.com [harryproa] <harryproa@yahoogroups.com.au> wrote:

Has anyone solved the autopilot problem with an HP?
As much as I'm a fan of tiller steering, handling two tillers far apart adds a challenge. Some purists may disagree, but I believe an autopilot is as mandatory as GPS on a bluewater sail, especially if single handed. 
Curious if anyone has managed to get a reversible autopilot working well with the daggerboard rudders on an HP?
Thinking two hydraulic arms for steering both ridders simultaneously connected to something like a seastar helm and autopilot pump... flip the heading sensor 180 after a shunt? And the pump wires? Or use an open center hydraulic valve to reverse the hydraulics (advantage of valve is having a neural gear to free the tilers for hand use) 
Have had hydraulic steering and autopilot pump before and it worked flawlessly.. so pretty confident in the reliability and maintenance of it.