This forum uses cookies
This forum makes use of cookies to store your login information if you are registered, and your last visit if you are not. Cookies are small text documents stored on your computer; the cookies set by this forum can only be used on this website and pose no security risk. Cookies on this forum also track the specific topics you have read and when you last read them. Please confirm whether you accept or reject these cookies being set.

A cookie will be stored in your browser regardless of choice to prevent you being asked this question again. You will be able to change your cookie settings at any time using the link in the footer.

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Project low cost autopilot
#11
Thank you very much Sean for your comments
In the first place we are no competition to your pypilot, we are not going to work with arduino or any arduino module engine.
We are designing a printed circuit board on which will go all components except the Gyro/compas.
On the other hand, unlike the arduino operating system that does not work in real time, we will work with FreeRTOS whose interrupt system allows you to control the PWM by performing the PID at 1000Hz. The programming language will not be with a language interpreted as Python but with C that is compiled. For this reason there will be no need to separate processes so as not to slow down the speed of execution.
So we would be delighted if you could join us and contribute your accumulated experience in your pypilot.
Regarding the power driver we want it to be able to be used also in linear actuators for ships of 20 Tm and that, despite the load that can suppose, they continue operating in the Safe Operating Area. Just look at the MOFET used by Raymarine in its autopilots (IRFZ48NS) whose Continuous Drain Current at 25º is 64 A, but at 100º it drops 45A. Its Pulsed Drain Current is 210A. This MOSFET has been mounted on both the S1 and the SP X-5. Between both models there are a few years of difference. The difference in our case, compared to your Pypilot, is that we have to design the circuit completely, since it is not a finished board like the Pypilot. But we see this rather as an advantage
I was very surprised by your comment about N2K and SeaTalk. OpenPlotter also works with N2K. Sailoog has announced that the new version of OP 2. will support the MCP2515 (CAN Controller for NMEA2000 or N2K) and your pypilot collects the OpenPlotter data. We are not going to do anything different. My boat has its sensors in the SeaTalk(old) system, the SeaTalk sensors work without problems. So it's easier for me to read these sensors than to mount new ones, e.g. open platform. That's why I want the autpilot to read the SeaTalk sensors. I also have a hot wire anemometer with SPI output for Raspi to read (https://www.youtube.com/watch?v=dl76Sz4mNWU). Now I have in development one of ultrasounds made, like yours, in 3D printing. The next step will be to develop a speed meter over water with ultrasounds that also measures depth.
On the other hand everyone has your competence in electronics and navigation. It is easier to buy sensors in N2K and mount them. That's why I think you should also read N2K.
We want to implement SignalK too. But we have to investigate if it means a decrease in processing speed. Of course using OpenPlotter would not be necessary.

I take this opportunity to answer Sailoog about the code. We do not yet have the code written.
What we want now is to obtain information in the forums to know what needs there may be that we have not thought. Then we'll do the hardware design and write the code. Finally we will test it.
Meanwhile, we are doing an analysis of the data provided by the IMU to do the data processing algorithm. Sean can probably help us here.
One of the things that interests us, is to know that pilots have the people in order to get a greater compatibility with systems that have already installed the navigators.
Reply
#12
Well we have already finished the board with the electronics, here you can see it:

[Image: Autopilot.jpg]

For the design of the plate, great care has been taken to avoid interference between the different elements mounted on the plate.  When digital and analog processes are combined, including switching systems with different frequencies, in order to avoid that these generated signals are mixed and transferred from one system to another, which would provoke interferences, altering its operation, it is necessary to build filters that prevent this undesired communication. Also in the design of the copper tracks it is necessary to be careful especially with the ground

The first concept has been to separate the inputs from the outputs. The inputs are where the data from the sensors of the boat (wind sensor, GPS, sensor rudder position mainly and other sensors) and the output that are the management of the actuator. Both elctrical motor, hydraulic and clutch.

For that reason in the left part of the plate these inputs have been placed and in the right the outputs for the actuator.

The SeaTalk and N2K converters are integrated in the plate. The conversion will be done in all standards. That is to say if SeaTalk enters there will be output in NMEA183 and NMEA2000 (or N2K). If you enter both SeaTalk and NEMEA2000 as NMEA183, the software will multiplex it establishing the corresponding filters and hierarchies to avoid conflicts and redundancies.  

Both the Bosch IMU BNO055 and the MPU9250 (the one used with OpenPlotter) can be mounted.
It will incorporate a WiFi module and a radiofrequency module.
The heart is the microprocessor STM32f105 or f405 with 8MHz clock.

The MosFet to handle the actuator can handle more than 100A, that is to say that the autopilot can be installed also in larger boats.

Now we are thinking about how we are going to connect the peripherals.
One of them is the IMU i.e. the one that generates the signals of compasses and gyroscopes and accelerometers.  With it you will be able to read the course, and the movements of the boat. This data will serve to compensate the electronic compass on the one hand and to optimize the rudder response according to wind and sea state by means of artificial intelligence.
This sensor should be placed at the point closest to the CG and in the centre line. The connection could be made either by cable (the safest) or by radio frequency.
Another peripheral thought, is a Joystick to control the rudder remotely. In this case the same thing can be done by cable (the safest) or by radio frequency. In this last case it could be done with batteries and thus place it where it is needed.
At the request of a user in another forum, we have included an extra entry of NMEA183, to be able to read MOB sentences for example of an AIS


Autopilot specifications:
-Power supply from 11 to 30 volts.
-PWM power output for arm motor: up to 30A.
-PWM output for clutch: up to 2A.
-3A relay output for motor stop (MOB)
-Internal connection of up to 3 I2c sensors. (Or LCD screen)
-1 full duplex channel NMEA0183 up to 38400 baud.
1 NMEA183 input channel up to 38400 baud
-Seatalk1. Bidirectional.
-CAN BUS. (NMEA2000) bidirectional.
-WiFi module. For firmware updates.
-2.4Ghz radio module for remote control.
-Internal FRAM memory for configuration and navigation data.
-Analog inputs for rudder sensor and joystick.
SPI input for other types of ensores.




Anyway, I'm not going to overwhelm you anymore. There are two ideas or questions to see what you think.
Reply
#13
"Anyway, I'm not going to overwhelm you anymore. There are two ideas or questions to see what you think."

You know a good engineer works to balance the design requirements and practical implementation. Drones use the mpu9250 extensively and they have to do minute rapid control loops to control a craft moving fast in 3 full dimensions. They work fine. You are talking about a craft that lumbers along slowly in comparison. Why would you need 1000hz control loops?

You seem not to be considering the simply adequate solution for the problem which is always the best engineering choice. Cheap and reliable.
Reply
#14
(2019-06-06, 05:27 PM)mike_kelly Wrote: "Anyway, I'm not going to overwhelm you anymore. There are two ideas or questions to see what you think."

You know a good engineer works to balance the design requirements and practical implementation. Drones use the mpu9250 extensively and they have to do minute rapid control loops to control a craft moving fast in 3 full dimensions. They work fine. You are talking about a craft that lumbers along slowly in comparison. Why would you need 1000hz control loops?

You seem not to be considering the simply adequate solution for the problem which is always the best engineering choice. Cheap and reliable.

We have developed and running a PID system with control at 1ms, although not for this application.
It is for another application that required the maximum precision and smoothness in the movement and at the same time enough power.
What do we start from this proven development and of course we will adapt the times.
Reply
#15
(2019-06-09, 12:36 PM)gypsylyon Wrote:
(2019-06-06, 05:27 PM)mike_kelly Wrote: "Anyway, I'm not going to overwhelm you anymore. There are two ideas or questions to see what you think."

You know a good engineer works to balance the design requirements and practical implementation. Drones use the mpu9250 extensively and they have to do minute rapid control loops to control a craft moving fast in 3 full dimensions. They work fine. You are talking about a craft that lumbers along slowly in comparison. Why would you need 1000hz control loops?

You seem not to be considering the simply adequate solution for the problem which is always the best engineering choice. Cheap and reliable.

We have developed and running a PID system with control at 1ms, although not for this application.
It is for another application that required the maximum precision and smoothness in the movement and at the same time enough power.
What do we start from this proven development and of course we will adapt the times.

Beautiful project. I put myself on the list. When it is ready?
Reply
#16
Hi, are there any news?
Reply
#17
Sounds very intressting,
I have now a garmin autopilot(2 years old). Hydraulic actuator.
The boat is a 42” KHSD catamaran
Don’t know if it will be up for the task of steering  in the south pacific at high speed.(10-20 kn).
The low cost autopilot should be able to use GPS-compass.
Preformance should be equal or bettet than B&G H5000 or nke.
Would this be possible.

Best Regards,
Peter Mannerstråle
Reply
#18
Thank you very much for your interest

We are finishing the software, regarding the configuration of the autopilot by the user.
Shortly we will begin with the tests in a sailboat.

(2019-10-12, 01:08 PM)skysurfer Wrote: Sounds very intressting,
I have now a garmin autopilot(2 years old). Hydraulic actuator.
The boat is a 42” KHSD catamaran
Don’t know if it will be up for the task of steering  in the south pacific at high speed.(10-20 kn).
The low cost autopilot should be able to use GPS-compass.
Preformance should be equal or bettet than B&G H5000 or nke.
Would this be possible.

Best Regards,
Peter Mannerstråle

The autopilot we are developing will be able to control both motor and hydraulic actuators.
The user can configure the type of boat and sailboat.
Regarding the sailboats in three categories sail slow turn, sail fast turn and sail multihull.
You can also adjust the Boat Normal Speed.
The autopilot uses a gyrocompass with GPS receiver for the three systems and accelerometers. It will also have the tables of magnetic declination according to position.

At the moment it is too early to compare it with other commercial autopilots, but our wish is that it will be as good or better than existing systems.
Reply
#19
Hi gypsylyon, I know that in the meantime you have made progress with your project, can you give me updates?  Huh Cool
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)