Trouble with boat thruster build

@will I know, I just wanted to make you aware of the challenge. I didn't even factor in vertical motion of the boat so now it's a 3D vector to deal with. @redaustin88 might have been onto something when he said gyro. I have an MPU6050 and it sports a 3 axis accelerometer and 3 axis gyroscope. It's a technological marvel at least to me as it is only 20mm x 15mm. I will try to find some specs to see if it's sensitive enough for 1 milli arc second per inch of movement.

@zander 

I'm curious why there's an allowable slop of 5 metres in the front point of the boat but only 50mm in the back.

I doubt that your rear propeller would be capable of maintaining position to that degree of accuracy in wind or strong current.

@will The GPS accuracy is that much. Can't do anything about it without lots of bucks. The rear is the end of a long lever so a small movement is the problem he is trying to address.

Assume for the moment the front does not move at all from it's GPS coords. Now his 14', 16', whatever length can pivot on the nose. A 4" movement is likely to be the point at which he wants to start correcting back to the start point.

Now, if the nose ALSO moves, the problem is waaaaaay harder. Now we need calculus to determine instantaneous angles. I don't do calculus anymore, but I know a hard problem when I see it.

@will The 6050 is not recommened for new builds so there is a replacement. Sparkfun has a unit with something called a DMP (Digital Motion Processor) that can do 9 axis solutions. It's $32.50 and OOS. I still have not found a spec that tells ME if the accuracy is there but my guess is maybe/probably. Try to source one of those units I mentioned and see the results.

Posted by: @zander

[snipt]Now, if the nose ALSO moves, the problem is waaaaaay harder. Now we need calculus to determine instantaneous angles. I don't do calculus anymore, but I know a hard problem when I see it.

That's why I thought of the compass. If it's given a heading, it will try to adjust the alignment of the boat to match that heading whether the nose is moving or not. It will always want to point the boat in the direction selected.

@will Not sensitive enough I suspect. However, I clicked on one thing and got another. Check this out and if you can figure out what LBS/DEGREE/Sec means we may have an answer.

https://invensense.tdk.com/smartmotion/

His OP that mentioned airplane might have been the right direction after all. I can be wrong and admit it.

@will @redaustin88 I think the place to start looking for a sensor is

https://www.sparkfun.com/pages/accel_gyro_guide

Read all 3 sections and pay attention to sensitivity, range etc. Also my previous link has products that may apply. Keep in mind how small a movement at the rear of the boat is enough to swing it about a fixed nose. It's a matter of a few inches.

@will @redaustin88 I stand corrected, for a 14ft boat, 4" of stern movement around a fixed bow is 0.02381 rad or 1.363 degrees. I would guess that level of precision is available for a hobbyist and likely a reasonable cost. Sparkfun seems to have a good handle on this subject and make sure to visit their 'Buying Guide' mentioned at the bottom for lots of info.

https://www.sparkfun.com/products/11028

https://www.sparkfun.com/pages/accel_gyro_guide

I did the calculation here https://www.calculator.net/triangle-calculator.html?vc=&vx=168&vy=168&va=&vz=4&vb=&angleunits=d&x=57&y=20

@zander FYI @redaustin88

I'm gonna back out of this discussion since it appears that I'm unaware of so many factors that I don't really feel I'm helping finding a solution here. Thanks for indulging me 🙂

Posted by: @redaustin88

head hold gyro

@redaustin88 FYI @robotbuilder FYI @will I stumbled on something I was researching for another fellow. You probably meant HEADING HOLD GYRO when you said head hold gyro. It's like calling a cat a dog.

YES that is what you need. Problem solved. I can't help any further though, maybe others can.

@zander 

I can't relate anything in your last post to the sketch or how he worded his explanation of the problem.

So does his sketch mean anything?

His Arduino output is "servo out" to a gyro, details unknown.
This "gyro" outputs to an actual servo clicking one of two micro switches which in turn are connected to nothing but apparently are the forward reverse control switches.
The "gyro" is also connected to another pot somehow that controls the pwm (speed).
The box with PWM on it connects the motor.
And where is the link to show the details of this "gyro" with its one input and two outputs.
Now if you can figure out what all that is about can you fix the sketch?

I am a visual thinker I find convoluted serial verbal descriptions less than useful.

He stated,
"It’s what I’ve used in the past but had to go to back of boat and change speed/ power and direction physically by hand or someone fishing in the rear got to keep control of the stern."

So what input is the person at the back using to control the stern that he want automated?

My assumption here is that there is an outboard motor at the back of the boat that can be used to control the stern. How does the controller know which way to turn the motor?

I thought that was done by the rudder.

@robotbuilder Not my project. Some of what you are saying is valid, some is not. Go back and read his first few posts. This is likely a total newbie that heard from a buddy he could do this but so far not enough info. I think it's doable but that is just life experience talking.

The key facts are an off the shelf gps controlled trolling motor holding position at the front but he needs to keep currents and wind from swinging his boat so he wants to use an old trolling motor at the back mounted parallel to the transom. Think of it as a thruster since it is variable speed and reversible he only needs one. Lot's of challenges but when I finally figured out that he was talking about a 'heading hold gyro' that he called a 'head hold' which meant nothing to I or @will . It is specifically designed to hold a heading using a multi axis setup, maybe 6 or 9, I don't know. He thinks he only needs the rudder (plane NOT boat) function. Probably true. AFAIK he has never coded or worked with an arduino, so now he is at step 1. Up till now we were all at step -something and going backwards. I have contributed as much as I can, so will let others now guide him into learning the arduino and then programming. 

@zander @will @redaustin88

Think I have it now.  The outboard motor would be fixed at 90 degrees to the boat and used as a thruster by running it forward or backward.  He wants the same automatic control used in a helicopter to bring the boat back to the right direction.

"Basically the heading is locked or held – thus the name heading hold gyro. This type of Control Feedback Loop is known as a PID control loop.

A strong crosswind in this case will not slowly turn the helicopter to point into the wind as it would with a yaw rate gyro due to the weathervane effect. The heading hold gyro will keep the heli pointed in the same direction until you give a tail rotor command to turn the helicopter to a new heading no matter how strong or gusty the cross wind is."

https://www.rchelicopterfun.com/heading-hold-gyro.html

@robotbuilder 

The situation is further complicated by the requirement for a GPS station required to constrain the nose of the boat to a fixed location. 

@will That is built in to the off the shelf bow thruster. I used to have one. Just set it and forget it.