PyPilot How difficult?

[onesea]

I had/ am seriously considering buying a:
Content Autopilot set - Autopilot

However after a week I have had no response, which is a bit of a concern.

having a spare tiller drive I was wondering how costly and difficult it is to make a PyPilot?
I have used Dos and Basic but have no more recent programming experience…

Out sailing yesterday we where considering what we wanted.

We would like a bulkhead mounted control not tiller mounted as dogs are often in the way back there, oh and accurate steering. Preferably with compass away from the cockpit over the years we have been amazed how many things are magnetic.

It seems our commercial options are not cheap…
Raymarine Evolution Tiller Pilot

 

[rowlock]

look at a pelagic.

 

[Tranona]

look at a pelagic.

They are not available in the UK and if imported from the States will cost at least as much if not more than the Raymarine.

 

[rowlock]

Tranonoa, not true.
Palagic currently about £1000 imported vat paid, Raymarine Evo equivalent £1500 plus. You are entitled to your own opinion but not your own facts.

 

[onesea]

Tranonoa, not true.
Palagic currently about £1000 imported vat paid, Raymarine Evo equivalent £1500 plus.

I make it just over £1000 That’s without cover or remote, which IMHO you would be foolish not to include, So nearing £1200.

The PcNautic version has what looks a better user interface, rudder feed back within the tiller ram. I understand will be nearer £850 vat paid. So not that much more than a budget tiller pilot about £500.

Its sad that non of the larger tiller pilot manufactures have not looked at redesigning what is now 30+ year old tech in cost effective tiller pilots.

For a start Even low end mobile phones have more accurate compass than a standard tiller pilot, that’s sad.

 

[GHA]

I make it just over £1000 That’s without cover or remote, which IMHO you would be foolish not to include, So nearing £1200.

The PcNautic version has what looks a better user interface, rudder feed back within the tiller ram. I understand will be nearer £850 vat paid. So not that much more than a budget tiller pilot about £500.

Its sad that non of the larger tiller pilot manufactures have not looked at redesigning what is now 30+ year old tech in cost effective tiller pilots.

For a start Even low end mobile phones have more accurate compass than a standard tiller pilot, that’s sad.

Quick update & back to the original question - Pypilot - pimoroni do a 9doff sensor for £15, just got a couple. Internally it gets the data 10 times a second.
ICM20948 9DoF Motion Sensor Breakout - Pimoroni
Was lucky enough to get one of the very rare Raspberry Pi's as well,2nd Pi4 as a spare.
Also you need a controller to do pump the current into the ram, yet to use in full on anger for a week but these seem to run very cool -
https://www.amazon.co.uk/gp/product/B07ZCS6YP9/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1

I installed pypilot by hand but think openplotter has already or very close to doing the work for you.

Then for compass data it's connect 4 jumper wires (the ones supplied with the motor controller work) and that's pretty much it. Pitch, roll, yaw, (as signalk data, not sure if there is an nmea0183 sentence for those) & mag/true compass output. It will run on a Rasp pi zero with very little power, offshore heard quite a few people turn the gains down so the course might meander a bit power power consumption is very low, wsy less than any commercial offering. Suspect the control side of the software is well ahead of commercial offerings as well, lower end of the budget anyway. But there are an awful lot of variables to weak if you want to fine tune it, getting close seems pretty easy though, tweak a few gains.

Getting the autopilot connected up is a bit more complex, need an arduino nano as a go between from the Pi to the motor controller. I've been fiddling with a PCB design from jlcpcb but needs a load more work, that would make the process so much easier & add little to the cost.

From longer distance reports it seems rock solid, avtivate a route on opencpn off it goes, many ways to control. Can steer to opencpn route, compass course, apparent wind or true wind (true is better for near dead downwind apparently. )

Anyone else here actually installed pypilot?

 

[ylop]

Interesting stuff, I've got a Pi4B with OpenCPN running on OpenPlotter and no autopilot at all. So I have been toying with going down the pypilot route. I hadn't realized the 9DoF was so cheap. I would be looking to control a motor directly coupled to the steering chain in my binnacle. Can I just use "any" motor or do I need to pay Jefa or similar a vast some of cash for what is probably a £100 motor. From what I read a rudder position sensor may be important. The other thing is having a way to control/stop it from the cockpit. I've seen some remote control type designs but I they seem a bit Heath Robinson for what I think should be the emergency stop?

 

[GHA]

Interesting stuff, I've got a Pi4B with OpenCPN running on OpenPlotter and no autopilot at all. So I have been toying with going down the pypilot route. I hadn't realized the 9DoF was so cheap. I would be looking to control a motor directly coupled to the steering chain in my binnacle. Can I just use "any" motor or do I need to pay Jefa or similar a vast some of cash for what is probably a £100 motor. From what I read a rudder position sensor may be important. The other thing is having a way to control/stop it from the cockpit. I've seen some remote control type designs but I they seem a bit Heath Robinson for what I think should be the emergency stop?

WIndscreeen wiper motors are popular as drives, think pretty much any brushed motor will work. Have a look on youtube, a few installs on there

Rudder feedback is probably best but the arduino is programmed to be able to measure current & use overcurrent as an end of travel indication.

Control can be over wifi with a phone or Pi, TV remote (not tried that yet) or wired. I got a little keypad just to have a try - needs 7 input pins though so might run out.
3x4 Matrix Membrane Keypad - Pimoroni

Lots in their forums > OpenMarine - Pypilot

Fair bit or work, but one guy ripped apart a st2000 & built a better one ??
OpenMarine - Pypilot

& it can run standalone >

 

[GHA]

good overview of available kit for RPI's here >
Hardwarebasis für AVnav und OpenPlotter – open-boat-projects.org

And more on Pypilot >
pyPilot – open-boat-projects.org

 

[dgadee]

On a Dehler 37cws there is no easy way to use a tiller device due to space issues etc. I wondered about those hydraulic steering controls from large RIBS. Didn't do anything about it and installed an evo wheel pilot which is useless.

 

[ylop]

On a Dehler 37cws there is no easy way to use a tiller device due to space issues etc. I wondered about those hydraulic steering controls from large RIBS. Didn't do anything about it and installed an evo wheel pilot which is useless.

I assume the reason that hydraulic steering has never been widely adopted on (at least sensible sized) yachts is that you get very little feedback on the rudder force. With cable or chains there is a direct feedback loop that tells you the sails are set badly etc.

 

[dgadee]

I assume the reason that hydraulic steering has never been widely adopted on (at least sensible sized) yachts is that you get very little feedback on the rudder force. With cable or chains there is a direct feedback loop that tells you the sails are set badly etc.

A friend has hydraulic steering but his yacht is not exactly designed for the cut and thrust of racing. However, my wheel steering is direct drive. What I was thinking of doing was attaching the hydraulic device to the top of the rudder post where an emergency rudder would go and it would work off the autopilot control, so feedback to human is not needed. Can't find a decent pic but this is from when I bought the boat:



The upright bit is the rudder post which goes into a bearing on the horizontal aluminium and a short squared section pops above that where the emergency rudder would go. There is no rudder post showing below decks - only a heavily laid up rudder tube.

That area was for the liferaft but I keep two gas bottles there.

 

[GHA]

On a Dehler 37cws there is no easy way to use a tiller device due to space issues etc. I wondered about those hydraulic steering controls from large RIBS. Didn't do anything about it and installed an evo wheel pilot which is useless.

A few treads on the openplotter forum (which is the main pypilot forum as well) about hydraulic control might be of interest.
Using a hydraulic gear pump to have a clutch

 

[dgadee]

This is the kind of hydraulic outboard control I had been thinking about, though power driven rather than from a steering wheel.

 

[ylop]

You'd need to engineer some signifcant mounts for the "piston" part and then an arm (a mini tiller) to attach to - the forces will be pretty high. Then you will need to engineer an electric motor to "helm unit" connection. Could you achieve the same thing just with a suitable motor and a chain?

 

[dgadee]

You'd need to engineer some signifcant mounts for the "piston" part and then an arm (a mini tiller) to attach to - the forces will be pretty high. Then you will need to engineer an electric motor to "helm unit" connection. Could you achieve the same thing just with a suitable motor and a chain?

Yes, I think that might work. Plenty of strength in that area - it takes the upper rudder bearing.

PS: didn't really think of this approach, but I can see that a motor attached to a based which is attached to the horizontal bearer turning via a toothed belt would work. I can't see any practical examples of this except windscreen motors. Any pointers to choice of motor etc.? Looks like an interesting project.

 

[onesea]

I was going to update this thread however I felt a bigger review would be more helpful.
PC Nautic Teller Pilot Review

 

[dgadee]

Good news. I have just tested mine under engine and headsail - not got down into the software innards but that's for the summer.