homemade autopilot

[bertho]

gents,
just a question for electronic fan...
is any "homemade" autopilot drawing on the net???
compas can be find easily, potentiometer for feebck too.. in betwen we just need to amplifie compas signal to send to actuating device... ? i'm not electronical specialist, far away, but look for me not "rocket science" to add adjustable temporization between compass signal and feedback rudder position.. , and send sufficient power for relay to drive on autopilot...?? any experience on the subject available??
bertho

[hoytedow]

The trick would be to make it precise enough to avoid zigzagging across the ocean.

[gonzo]

It is a complicated electronic equipment that has to work in salt water. You can use any hall effect sensor to make it work. Then you have design and build a circuit to adjust for sea state and gain. The compass has to work at different angles of heel, pitch and also pounding and vobration. That requires damping and either gimballing or a compass that works at any angle.

[catsketcher]

I would really love to do my own autopilot. I have been bugging my electronics mad son to design me one but he says it is pretty hard.

I don't like the response time or multihulls of the typical autopilots. I would also like to get a proper sized motor without paying huge dollars.

So

What about using cheap accelerometers (I phone - Segway) and have these providing info to a box that takes inputs from a GPS. This does away with the fluxgate compass. I would especially love to be able to fine tune the whole process by using my laptop and help the box learn. For instance it would be fab to be able to tell the box (via laptop) that the boat is on port gybe. Then when the nose dips (accelerometer) down for a long period of time and the speed builds up GPS ( the boat starts surging on a wave) that the pilot bears away. This is what any helmsman does automatically.

Any ideas?

cheers

Phil

[farjoe]

Originally Posted by catsketcher

What about using cheap accelerometers (I phone - Segway) and have these providing info to a box that takes inputs from a GPS. This does away with the fluxgate compass.
Phil

I am actually in the process of doing something similar. i have decoded the Track info from the NMEA string send by the GPS using a micro controller and written a simple algorithm which drives 2 pins according to the deviation from the chosen track.

My plan is to set this up on the boat and see how it reacts and then go for a more sophisticated algorithm later.

I am stuck at building a proper housing for the motor assembly and fitting it to the tiller of my cat.

Any ideas?

[TerryKing]

Take a look at what the model boat (and model airplane!) serious hobbyists are doing. And then, for complexity, what a few have done with model helicopters.

I don't think the guts are very expensive.

The ocean is the same ocean. The dynamics are similar but different (they have to have faster response than a large boat). The servo has to be bigger.
But most is the same.

Update: There's real stuff out there - Look at this:
http://paparazzi.enac.fr/wiki/Tiny_v2

A few pointers:
http://www.sciencenewsforkids.org/ar...7/Feature1.asp
http://news.stanford.edu/pr/96/960116gpsplane.html
http://wreilert.stud.hive.no/hopfe60...ds/rapport.pdf
http://www.gpss.tripoduk.com/autop.htm

Here's an open-source project for one:
http://autopilot.sourceforge.net/faq.html

Another more current project:
http://paparazzi.enac.fr/wiki/Main_Page

And these guys are doing serious stuff:
http://diydrones.com/
See this:
http://diydrones.com/profiles/blogs/...BlogPost:35640

Here's an idea: A 'smart' cellphone has all almost all you need inside for about $200. And you can send it a text message "come home"..

I'm living at new big University (King Abdullah University of Science and Technology http://www.kaust.edu.sa ) on the Red sea, with a long canal between housing areas. I've been thinking of doing a project with high school kids with a model boat that can auto-navigate the canal and the harbor.

I'm sure this is possible to do...

[Manie B]

I had a look at this type of thing and in South Africa it would just be too expensive and then you have to worry about reliability
kinda re-inventing the wheel

[gonzo]

You can buy a Tillerpilot for about $250.00 US dollars. The gears are not too sturdy, but if you connect it to a servo it will do very little work.

[CDK]

Originally Posted by bertho

compas can be find easily, potentiometer for feebck too.. in betwen we just need to amplifie compas signal to send to actuating device... ? i'm not electronical specialist, far away, but look for me not "rocket science" to add adjustable temporization between compass signal and feedback rudder position.. , and send sufficient power for relay to drive on autopilot...?? any experience on the subject available??
bertho

Yes, I made one several years ago, but the results were disappointing. My objective was to keep my boat, with twin stern drives, on course during displacement in choppy seas, so I didn't need to hold the steering wheel for hours.
I used the compass from an old Autohelm unit. That is the critical part; it hardly looks like a compass, just an oil filled cavity in a clear plastic cylinder with a circuit board containing some simple opto-electronics that determines the deviation from course and the direction. The servo amplifier and output stage are child's play for an electronics expert.

It should work reasonable well on a sailing boat with a keel, but in my case steering was very nervous due to the many over-corrections.

With today's technology it is much easier.
Use a fluxgate compass, a linear actuator and something with a microprocessor. Write a simple program that ignores the static from waves and considers the response time of the hull.
It is still on my "to do" list, but not on top.

[yipster]

that girocompas is essential, rc helicopters nowadays have it and use 2 channels instead of 5.
had an autohelm on dual engine boat years ago too, girocompas down near CG
between wheel and dash the motor and controls, on/off, 3 different speed actuators when going fast
and a independant stay on compas course button, quite a nice system i hooked to gps and old laptop with nav soft
could set course between islands in advance, worked well but never used it much
till lever with all the buttons came off after to much sun and never repaired it
NMEA connections on serial? pc port was it? had more trouble hooking the gps into the system
ah, remeber now, was one of the first handheld and NMEA out didnt work i found out
whole batch went back to usa and got a new one but it was a headache to find
think a toyshop gyro compas and motor may beat the was was it 1500 i paid
wannahaves it are, nice, but you probably wont use it much, need an eye at the helm anyway

[farjoe]

Which characteristics of a girocompass or a fluxgate make either essential for an autopilot?

Why isn't the track data coming out of a modern GPS sufficient? It may not be correct to compare older GPS units with modern ones. The one i like most is half the size of a computer mouse, totally water proof, very low current comsumption and can almost detect satellites from within the house.

It seems to me that with a simple microcontroller the variation due to waves can be averaged out. For me the challenge seems to be more in determining the actual angle of the rudder than any other parameter since many units I have seen do not have direct feedback of rudder angle.

of course this is all theory and practice may teach me otherwise.

[Fanie]

This is not difficult to do at all.

You need two things for bearing - a wind vein and a compass module. One would be the master and the other one a reference so that if the reference vary more than a preset amount you get an indication or alarm.

If you follow the gps and the wind direction turn you must be notified, otherwise you may be sailing back instead of ahead

If you follow the wind and it again turns you may veer of course some.

The drive for controlling the rudder can be made quite easily. A rack and pinnion can be made from a wiper motor (you get big ones ie for busses as well) and the rack and pinnion from a gate motor available in all plastic too. You could even use a 12V gate motor.

This should be mounted inside the hull. Then a push pull cable like they use for boat steerings takes it to the great outdoors and onto the rudder.

On manual control you can use a joystick to steer.

[CDK]

Originally Posted by farjoe

Which characteristics of a girocompass or a fluxgate make either essential for an autopilot?

Why isn't the track data coming out of a modern GPS sufficient? It may not be correct to compare older GPS units with modern ones. The one i like most is half the size of a computer mouse, totally water proof, very low current comsumption and can almost detect satellites from within the house.

I would prefer a fluxgate because it costs almost nothing and can measure the hull position angle. GPS gives the traveling direction with the resolution of the GPS grid, good for sailing but not to keep a motorboat on track.
A girocompass is something from the past that need not be discussed here.

[Fanie]

A girocompass is something from the past that need not be discussed here.

Carefull CDK The new ones are very fancy. The 2 axle ones I agree with, to use them they should remain horizontal. With the three axle ones you can compensate for the boat heeling, and you hence get a very accurate and stable output. You do however need a processor to interface them.

The biggest advantage these have over GPS is that you can always see dierection, with a gps you have to be moving...

Have a look at the Honeywell HMC5843

[CDK]

The Honeywell HMC5843 is 3-axis fluxgate sensor.
To me (must be my age) a girocompass is a fast spinning flywheel on a hinged platform. The last serious application I am aware of was the LN3 inertial navigator used in the Lockheed F-104: the error after one hour was more that 20 miles!