homemade autopilot

Discussion in 'OnBoard Electronics & Controls' started by bertho, Nov 8, 2009.

  1. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    Pixhawk

    Is anyone attempting to use a PixHawk to do marine autopilot control? I'm involved in a land based autonomous vehicle system development right now and the PixHawk was going to be used for the vehicle guidance/steering system but the PixHawk seems to have a lot of issues right now. Anyone else seeing issues with the PixHawk? If so what are your plans.. to wait out the fixes or have you developed a Plan B. If we can't find an alternative, the project may go on the shelf for a while. There is no money to develop an alternative.
     
  2. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    Work continues ....

    After working on my "new" boat this last summer to get it into good sailing condition I am back working on a better autopilot solution. I'm going to publish all of the work I have done so far on my blog site soon including wiring diagrams and programs if someone wants to get a jump start on what works and what does not. My last design used a Click PLC available from Automation Direct. The actuator was an off the shelf electric unit that worked very well. On my "new" boat there is an Autohelm 4000 wheel pilot already installed but it is getting up there in age so I am working on a second design as a backup. Initially it will use the existing PLC based control (although I will likely change from a Click PLC to a newer Siemens PLC) and since my last design was for a tillerpilot type autopilot and my "new" boat has a wheel, the two are not immediately compatible. I wanted to go with an electric actuator for the "new" boat which would turn the quadrant below decks, but I have not been able to find a decent actuator that can be clutched in and out (for manual steering) that does not cost a small fortune. The old Autohelm Type 1/Raytheon Type 1 is the best I can find. Used, in uncertain condition, they can purchased off Ebay for several hundred dollars typically. New they are $1500-1800. And while they are durable, that is a lot of cash to put into an older boat for an electric actuator. So I have decided to go hydraulic. Type 1 reversing hydraulic pumps are fairly common and steering cylinders are also common. For a first pass I picked up two used type 1 Autohelm pumps and for a steering cylinder I purchased what looks like a barely used outboard cylinder made by Seastar that has the correct dimensions. All of that set me back about $450. I chose the outboard steer cylinder for one reason, parts are widely available. If I need to reseal it, the parts kits are sold on Ebay. Plus it is all stainless and power coated alum which should be long lasting underdeck near the quadrant.
     
    1 person likes this.
  3. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    My Autopilot design, program, wiring online

    Here is my blog where I posted my Autopilot design.

    http://www.dc9.com/blog

    I need to update the electrical prints (not complete) and add a few photos but the PLC program code is current and so is the screen configuration. This has been successfully used on my Ericson 29 sailboat. There is no rate gyro in this design. The use of a rate gyro seems to be required when running downwind in a sailboat. Very few autopilots can do this correctly. Factor in that IOR type sailboat hulls (canoe hulls with fin keels) are basically unstable when running downwind with following seas and I think the chances of getting an autopilot to effectively work downwind with an IOR type hull is nil. At one time I wanted to try and integrate a rate gyro into this design but I'm not going there. I don't think it is worth my time.

    My next exercise is to update this control with a newer PLC that has a built in webserver and get rid of the dedicated screen. With the webserver I can tweak the autopilot tuning from a tablet. I'm also going to use this same control with a hydraulic actuator.

    The cost to get started with this setup is not great.
    This PLC costs $129. The added I/O card another $40. A DC to DC converter about $30. The electric actuator about $100. The compass.. well that depends on what you use, but less than $100. A few relays $20. If you want to go with a cabled pendant (recommended) you can add another $100 for the waterproof joystick, box, and pushbuttons. What you have when you are done is an autopilot that you can service and update.

    The nice thing about using this basic design is that it is adaptable to an electric actuator, a hydraulic actuator, or even a stepper motor drive (think wheel pilot - given a different PLC).

    Regarding the use of this design with a powerboat; This should work fine. In fact I have realized that powerboat autopilot design is really simple compared to a sailboat.
     
  4. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    Hi Guys...

    I added the ability to leave comments on that blog link I sent regarding the autopilot design. I still need to add some pictures to that build. But at least now you can comment on the design.. or tell me if I left something out.. etc

    I'm getting a bunch of hits on that blog so apparently there is still some interest in homemade autopilots... but this message thread is really quiet.. :-(

    Dave
     
  5. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

  6. pdwiley
    Joined: Jun 2008
    Posts: 1,004
    Likes: 86, Points: 48, Legacy Rep: 933
    Location: Hobart

    pdwiley Senior Member

    I'm interested, just don't have anything to add at the moment. I've checked out the blog, need to take a look at the code etc. I'm not a PLC person so not sure what will make sense to me. C, C++ or Java, fine.

    PDW
     
  7. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    Thanks for the note... I was starting to wonder if anyone had any interest in this stuff any longer.

    The Automation Direct software is a free download at:
    http://support.automationdirect.com/products/clickplcs.html
    It is not a huge download at only 37 megs and it runs on everything up to Windows 8.1.
    The logic I used could be translated without much difficulty to an Arduino or something like, but the robust case of the PLC and onboard 24 volt I/O of the PLC makes things a lot simpler.
     
  8. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    I was thinking about changing out this simple Click PLC for a Siemens PLC that supports a webpage, but now I am reconsidering. The Click PLC is just soooo cheap and I already have it and it is programmed.
    I'm thinking Android device for the autopilot operator panel connected to this PLC via a Bluetooth to RS232/RS485 converter. With some work I think the compass in the Android device could be used in place of an external compass saving $$. I have a pile of Android devices looking for a purpose. I'm already going to use a tablet at the helm for navigation on my sailboat and it is going to be stationary, in a mount. So I should be able to send the compass heading to the PLC via Bluetooth, and read data back from the PLC via Bluetooth to monitor the autopilot. Running the autopilot app and the navigation app at the same time should not be a problem. (I think..) The tablet I was going to use for Navigation is a 1.5 gig hz dual core tablet so you would think that would have plenty of HP to run two apps. I found a bluetooth to RS232 converter device for just over $50. I just looked at the Android development package and it seems like they have simplified that quite a bit. By doing this I would get rid of the Automation Direct screen (-$160), add in a Bluetooth converter (+50), and hopefully get rid of the standalone fluxgate compass (-$100 at least) so the net results is a $210 savings if you have a tablet laying around you can use. So the total system could consist of a PLC based control box with a bluetooth interface to handle the I/O and loop controls, a Tablet/Smartphone, and an actuator/rudder feedback device. This seems like the simplest solution without any custom boards..nonstandard equipment. Feel free to shoot holes in this idea.... I keep wondering if a Pixhawk could be used for an autopilot. Looks like they are planning on adding Bluetooth.
    I just noticed this thread, but I can't find anyone who has actually tried tweaking a Pixhawk to control a boat. http://www.cruisersforum.com/forums/f2/diy-autopilot-118491-2.html
    I was involved in a project that used the Pixhawk for a rover application (wheeled autonomous vehicle) and it turned out to be very difficult to modify the Pixhawk code. The developer told me that a lot of the Pixhawk code is not documented.
     
    1 person likes this.
  9. AndySGray
    Joined: Jun 2014
    Posts: 296
    Likes: 13, Points: 18, Legacy Rep: 91
    Location: Cayman

    AndySGray Senior Member

    Thanks Dave, great info, much appreciated

    :)
     
  10. Rabadak
    Joined: May 2015
    Posts: 4
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Portugal

    Rabadak New Member

    My Autopilot project

    Hi all,

    Just started my Autopilot project.
    The idea starts after I attached a linear actuator to my outboard with a remote control (April/2015).
    So, if you want give a look, please see the PDF.

    Regards,

    Rabadak
     

    Attached Files:

  11. Rabadak
    Joined: May 2015
    Posts: 4
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Portugal

    Rabadak New Member

  12. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    Autopilot design continues...

    Well its 2017 and I am still kicking around better solutions for this.

    The old wheel pilot on my new (old) boat works ok so I don't HAVE to get an autopilot working at the moment. But I still want to get a below decks hydraulic autopilot setup and running if for no other than than long term reliability. My Autohelm wheel pilot has a new motor but the old gearbox and old control box has not been changed at all.

    There are some HMI Android apps on Google Play now that work very well with known protocols for some PLCs and Microcontrollers.

    For testing purposes I have a "HMI Droid" app ($7.00 from Google Play) running on a Nexus 7 tablet connected via Wifi to a Siemens S7-1200 PLC. It works really well. Basically you can put buttons and displays on a standard Android device via an editor on a PC and load the screens into your Android device via a standard USB to MicroUSB cable. The screen talks via the protocol selected to the controller. I'm using the Siemens ISO on TCP protocol. The Siemens S7-1200's have an Ethernet programming port and can host Siemens comm protocols and Modbus TCP as standard. Using the Siemens comm protocols negates any need for comm programming inside the PLC. For the PLC Wifi connection I am using a really cheap TP Link Wifi Router that I bought for about $15. Its smaller than a pack of playing cards. Wifi goes right through a fiberglass boat so no issues there.

    This same HMI Droid app can talk Modbus RTU (serial) and TCP to microcontrollers as well. (Think Arduino and Beagle Board Black) So things are coming together. I have a growing pile of old smart phones and tablets. The Nexus 7 is 4 years old. But I saw a $59 8" Tablet at Sam's Club that looked pretty good this last week as well. Samsung has a waterproof 8" Tablet for about $170 that is apparently very readable in sunlight! So if well done, a tablet and app like HMI Droid along with a suitable PLC or micro controller could probably replace a lot of instruments in the cockpit.

    An old smart phone could make a nice remote control as well. I would keep power on the Tablet so the screen can stay on all of the time. They draw so little power that it would be almost insignificant. Another tablet app such as MX Mariner could also be used as a chart plotter all running on the same tablet!
     
  13. Rabadak
    Joined: May 2015
    Posts: 4
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: Portugal

    Rabadak New Member

    Autopilot design continues...and complicating

    With all respect Dave. I think you are complicating a simple thing.
    With all described components, I think you are in a maze, running and running but no light at the end of the tunnel.
    Sorry, it's just a thought...
    A solution has to be simple or is not valid (just another thought...)

    I wish you get your goal, but I think it's not the best way.
    :confused:
     
  14. Dave911
    Joined: Aug 2011
    Posts: 70
    Likes: 2, Points: 0, Legacy Rep: 16
    Location: Indiana

    Dave911 Junior Member

    Hi Radabak,

    >>With all respect Dave. I think you are complicating a simple thing.<<

    I don't think so.

    I've already made one autopilot and tested it. I documented it in this thread about two years ago. See my previous posts in this same message thread. I had it running a little over two years ago. You can also see it on my blog at DC9.com.

    But I'd like to improve the design. I'd like to refine the hardware and interface and go with a full hydraulic design. The electric actuator I used is very similar to the one you used. And it was ok, but not for running 24x7 and in rough weather. I want to put everything below deck, out of the weather.

    Dave
     

  15. AndySGray
    Joined: Jun 2014
    Posts: 296
    Likes: 13, Points: 18, Legacy Rep: 91
    Location: Cayman

    AndySGray Senior Member

    I'm with Dave on this, Rabadak. There is a lot of value in using off the shelf and modular components and you are leveraging massive R&D budgets.

    Don't confuse value and price;- A samsung display for a mobile phone or tablet may pass through the checkout at a few dollars, but they are aiming at a global market in the tens of millions of units.
    Furuno, Raymarine and Garmin are aiming at a market which is a minuscule fraction of that size and indeed they will be doing the same as dave, If you take apart your very expensive Autopilot screen module the display would be made by Toshiba, Sony or their ilk.

    The reason you can now get a handheld GPS chartplotter for small change is exactly because of components developed for other applications;- better screens, lower powered processors, more sensitive gps chips and so forth.

    Furthermore, simply because you are seeing all the links in the chain, doesn't make it more complex than a commercial unit, it is just that the commercial unit bundles many of those links together in a 'black box' module, giving the illusion of simplicity...
     
Loading...
Similar Threads
  1. Deering
    Replies:
    2
    Views:
    1,758
  2. Bumbleguy
    Replies:
    3
    Views:
    3,284
  3. missinginaction
    Replies:
    1
    Views:
    4,937
  4. sv alexandra
    Replies:
    2
    Views:
    3,445
  5. missinginaction
    Replies:
    1
    Views:
    1,922
  6. missinginaction
    Replies:
    2
    Views:
    2,049
  7. Jdawg
    Replies:
    9
    Views:
    2,579
  8. Tony.R
    Replies:
    0
    Views:
    2,077
  9. Steve W
    Replies:
    0
    Views:
    1,745
  10. cal_d_44
    Replies:
    0
    Views:
    2,081
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.