Okay, so I build a crazy self-steering contraption...
and it actually works!!!

currently self steering a beam-reach from Bay of Islands towards Cavali Islands.

It's enspired by a "servo pendulum" design, but only uses junk you probably already have lying around. So far I have spent $30. $20 on rope and $10 on a pulley. (although I already had a few of those in my junk drawer)

So the design is a horizontal windvane, made from coreflute (aka coroplast) a door hinge and some bolts as a counter weight.

Then it goes to a little servorudder at the end of what was the boat-hook (which was previously a broom, before I lost the original boathook). Control lines go from the windvane to the servorudder, and then heavy lines go from the shaft of the servo to the tiller (and thus the boat's rudder)

The lines cross over, so that the gear tends to correct the boats movement - if it didn't cross over, movements due to the waves would get eccentuated, which could cause eratic steering (in theory, maybe I can change this)

It's actually pretty impressive that this cobbled together thing even works. And now that I've seen it in action, I bet I can improve it a bit too. The servorudder really isn't moving very much, and I can see it stalling sometimes. A longer better foil would improve this. The windvane is also held in place only with friction, and doesn't have much leverage over the servorudder.

There is also quite a bit of slop because I've used ropes for all the linkages, and adjusted them with trucker's hitches (a knot where you can add tension, which truckers used before they invented ratchet straps)

It also worked on the quarter (45 degrees off downwind) I just adjusted the windvane and then the sails! to windward works too, but that is the easy one - just tieing the tiller off with a rope works to windward.

TODO: try dead downwind, stronger wind, lighter wind.

I found it surprisingly difficult to google for a basic explanation. It took me a couple of youtube videos to figure out what this contraption actually does in practice, I think.

So in a nutshell you can set a heading for the boat in relation to the wind direction, therefore using the wind as a kind of compass. Is that right?

@ktorn even better, using the wind as a helmsman. A compas just tells you what direction you are heading, the helmsman keeps you going in that direction.

The wikipedia page is pretty good, particularily the history section: https://en.wikipedia.org/wiki/Self-steering_gear

Also this article is good, http://www.faymarine.com/plansite/fay_marine_yacht_pl.htm (although unfortunately it complete avoids discussing the most interesting part, negative feedback to prevent oscillation) It covers the 4 most common configurations - but note that my trailing servo style isn't described - I have never seen or heard of anyone using this style, so it might be my invention!

here is a close-up of the servorudder

The webbing strap plus wire and bolts seems to be rigid enough. Once the design stablizes I'll make a proper waterproof one sealed with epoxy, of course.

And I modified everything around the control lines and the power lines, making it easier to adjust (without having to stand on the duckboard) and reducing friction. Also I switched the wind vane pole for an old paddle (which I found washed up on the beach one time and kept, knowing it would be useful eventually) It has a handy mechanism for locking at a particular angle (twist the black grip thing) and also provides a disengage mechanism: slide it shorter, removing tention on the servo!

Easy control line adjustments:

vane pole moved outboard (provides a straight lead for the control lines, reducing friction)

and moved power line pullies, so they are at better angles, less friction and clearing away from other things stored on the rails (LPG tank, outboard)

This version is great! before it worked, but I didn't really understand why. Now it clearly works as I understand it to. Sailing on it now, back towards Bay of Islands.

Although, to windward im not really sure how much it's doing, because just tying off the tiller with a rope works pretty well. I think it's working slightly better that a rope does.

Excellent, now you can't use your sneaky pulling dingy out of the water unsportsmanship trickery to cheat at races anymore

it's working pretty well right now, gotta resist temptation to post to ssb don't wanna get distracted and hit something! I need one of @substack's cybernomad rigs

Yesterday I was joined by Doug and we sailed from Opua in bay of islands to Oneroa on Waiheke island, in one go, in 34.5 hours.

At the start there was no wind, so we motored and took turns steering, until we got around cape Brett (that is the big right-hand turn) at that point, we had a very light tail wind so we raised the spinnaker to make the most of it. Normally the spinnaker is very finnicky to steer for, it likes to have the wind from just the right angle. With a human steering their mind wanders, and they let the boat go off course and the sail flaps and reminds them, but the mechanical self-steering keep course way way better! It sailed us down wind for hours, no problems! We made pretty great time in that stretch, around 5 knots.

But then the wind changed, to a westerly, and we had to switch to the genoa (large jib) again. It was pretty much a close reach or windward the rest of the way. We decided to sail through the night and took two hour watches. Two hours passed pretty quickly and was good enough sleep (and we were tired enough to fall asleep pretty quickly!). On watch there wasn't much to do but make sure we didn't hit anything, just hangout pretty much (and occasionally adjust the steering or sails)

Then the next day once we got past Kawau the wind picked up and we eventually reefed down to number 2 (medium) jib and double reefed main sail, bumping along quite quickly, and pulled into oneroa with still time to go spearfishing for dinner!

The self-steering struggles most when the wind is on the beam (coming directly from the side) I guess that is where the most weather helm is. I think if the servorudder was a little bit bigger it would handle that (also, I can see it stalling sometimes, because a wall of bubbles flys off the leading edge when it's trying to resist weather helm, but it's just a flat piece of 5mm plywood, so a proper foil shape would work much better)

Also, in Opua I upgraded the control line to 3mm dynema (which is low stretch and low friction) and better pulleys. The pulleys said something about kite surfing, they where $16 a pair, which was $6 more than the (shit) galvanized double pulley I had previously, but worth it. I still have some dynema left over. The total expendature is now $60.

and added this plastic smoother (a pen) which holds the line away from the edge of the aluminium pipe.

sailing is now just hanging out in the shady part of the deck

haha @neftaly it wasn't me, but last time I was in these parts I met someone matching that description - John in "Penultimate" (name not written on boat, though)

I have large solar panel on pullpit, duckboard, and red plastic kayak on deck. CLEO written on bow and transom.

I made some improvements, and now it works on the beam, good enough anyway.

I made a much better fin - it's longer, and more fin shaped. It's sealed with epoxy. There is a hole drilled from the top, with a bolt into it. The green rope is epoxied on to reinforce what I expect is the most stressed part. Okay so I spent another $5 on rope but only used a tiny part on this. Shown next to the previous fin. Also I brought the control lines in closer to the pivot so it is more responsive.

Because the fin is more powerful, it broke pivot at the other end of the pole, so I had to make that better two. The second attempt worked.

It took a few tries to get this to all fit together right, but it works better now! I sailed back from mahurangi with 15-20 knots on the beam, to top of motutapu, no problem. Then today back to auckland with 25 knots to windward. Can tack by just turning the wind vane then, handling the sheets! (just don't trip over the power lines to the tiller) Maybe I can rig up a remote adjustment to the wind vane.

Oh yeah, the windvane wobbled a lot in 25 knots, maybe need a way to reef it?

@c3 oh regards the board steering, note that old ships (vikings, etc) put an oar like rudder over the side, rather than the stern, on the lee side, I think. Probably somewhat better for non-exausting steering. Although, with the extra length behind the boat, the rudder has more leverage.

Starboard = "Steorbord" = "Side you steer from" = most people are right handed so put it on the right = when you're heeled over with your steering board out of the water your steering sucks so you have PRIORITY over sailing vessels on other points of sail = GREEN = I AM RIGHT GET OUT OF MY WAY

@dangerousbeans that makes sense! some cursory searching backs up your story. also, at one point left was "larbord" which was nice because it rhymed with starboard but that caused problems because you have to yell it into the wind and they might hear the other one, so they changed it to "port". Remember "there is a little port left in the bottle" (also note, that port wine is red and so is the port side light)

Ahh that's a good way of remembering

I always went with "starboard" has more "rrrrrrrrr" noise in it which is the same as "Right"
And right is the same as correct or green.

@nrkn I still have to sing the alphabet song in my head to work out letter ordering :C

Having this week made a trip from Auckland, to Waiheke, to Great Barrier, to Kawau and now night-sailing back into auckland, it's a good time to update this thread on design progress.

The was noticing that the steering was getting a bit sloppy, and indeed the control fin would not just yaw but also roll. Examining the bit where the control fin pivots on the lever end - it was badly worn!

The bearing is a stainless steel bolt directly on a pretty thin walled aluminium tube, and aluminium is pretty soft so it's not surprising. I tried fixing this by rotating it and drilling another hole, but later realized that the original drill bit had broken at some point, and the new hole was oversized, so same problem. I found a hard(ish) plastic tube that fit snugly between hole and bolt... but the pressure between the bolt and the tube cut it, so I tried filling the end with an epoxy plug.

I used some rubbish as a plug, then poured epoxy into the end.

When that went hard, I had a working self steering again!
I was sailing with that from Kawau to Waiheke, and everything was going pretty well, then dolphins came (at night again - saw dolphins 3 times on this trip!) and then suddenly I noticed that the control fin was rolled way over, then I saw it barrel roll right around the lever!

I pulled it in and the bolt had snapped. Did a dolphin break it?
Anyway, I managed to get the snapped bolt out by cutting a groove in with the hacksaw and turning it with a screw driver.

Clearly, the control fin pivot is the most highly stressed part!
It's a fin about 500mm tall, canterlevered off a 6mm bolt going through 25mm tube. If that was a 50mm tube, it would be a lot better, with some sort of sturdy bearing... Solid plastic, and a thicker bolt... The level is also under torsion, so something with a higher cross sectional area will make that stiffer.

I want to not have to fix something every 50 miles!

@dominic this may sound like a silly question, and maybe was already discussed, but what's the feasibility of using a small 3D printer in the boat? The ability to quickly produce DIY plastic bits seems like a major feature when you're in the middle of the sea and don't have access to anything else.