GPAPB and tacking in wind mode

[rmvdmrl]

Hi Sean, 
 
I have recently (have been away for a while) upgraded TP to the 21122019 and noticed that the GPAPB was switching to compass mode again (and code on GIT is still the same) so I've made a Pull request on GIT for that. I did test it on my TP and seemed to work.
 
I also noticed that tacking in wind mode and ap active results in a crash and restart (connecting to server)  I hope you can look in to that to me it seems something must be in line 124 (tacking) but can’t see what’s wrong.
The code for tacking seems complicated to me with the turning rate and the PD filter. Hope if others already have some experience with tacking on compass mode they are willing to share this here. I have not been able to try this on the open (Well I have been trying but i didn’t calibrate the IMU on the new images and got no good reading) and tacking in the harbor (and with a non calibrated IMU) does turn the rudder a bit (5 degrees or so), halts and gives an occasional puls. I guess there is no turning so no turning rate etc.
 
Hope to try it out for myself soon.

Greetings Rob 

[seandepagnier]

the tacking support is experimental and not well tested but it would be great to hear about more possible strategies that could be used.

I am working on a lot of improvements to pypilot and it's unfortunate tacking has not developed better.

[ironman]

I must say I've never missed it; the +/- 100 buttons are a perfect KISS alternative for me.

[seandepagnier]

Tacking depends on the type of boat and the conditions. So in many cases +- 100 degrees or so works, but in cases with less wind, or more waves, and depending on the boat it will fail to tack at times and/or overshoot. Integrating waterspeed as well as heel as inputs is needed to compensate.

So there is no one clear solution to tacking, and other than self tending rigs, requires coordination with manual sheeting as sometimes the head sail must back wind to complete a successful tack.

[rmvdmrl]

Thanks for the reply's, 

I'm now starting to understand how Sean made the Tacking and why. 
I have tried tacking with the compass calibrated but it didn't turn the boat with the set tacking angle of 100 degrees. At most it did about 10 degrees (with default settings treshold 50% and turnrate 20 degrees/s). I will play with the settings the next couple off days (turnrate and treshold) to see if it makes a difference. 

It would be good if tacking with turnrate and treshold was working the way Sean designed it. 

On the other hand I think it would be not that difficult to implement that if the user sets turning rate to '0' the TP would just do a "simple" course change with the set tacking angle. This way it is also possible to do a simple tacking or +/-100 (tackangle) from the buttons on TP without OpenCPN/OpenPlotter.

Another option to make tacking go faster could be to turn the rudder about 30 degrees in the tacking direction and then set the new course.

I will let you know the results.

Regards Rob

[rmvdmrl]

Well I have tried different settings for turnrate tackangle and tackrate. It does not realy matter what I set the result is the same: the boat does not tack. It does try to do something, it moves  the rudder about half a degree or so every couple of seconds.  This could result in a real slow tack that would take about 20 or 30 minutes I did not have enough clearance or time to Let it go complete. Perhaps something wrong without the calculation for tackrate or so. I have not heard anyone that was able to do a tack with this procedure so I quess it Just has some problems I hope Sean can find some time to look in to this.

Regards Rob

[seandepagnier]

Sorry for this. Definitely something is not working right. The motor is supposed to just move at the tack speed until the boat's heading reaches the tack threshold before switching back to normal autopilot mode on the new tack. This is simplified and obviously isn't the best strategy since tacking in different sea states need completely different settings for this.

With rudder feedback tacking could probably be improved, and it would be interesting to use rudder angle here. The rudder can move to a known angle and wait until the tack threshold is reached. Without rudder feedback this also may work better, by moving the rudder a constant amount and waiting.

[ddelorme]

I am having issues with pypilot seeing wind data. my wind data is coming into port 10110 as nmea data. Opencpn can see it? Do i need to send it to 20220?

[seandepagnier]

yes, you need to send it to tcp 20220

[johnm]

Sorry for this.   Definitely something is not working right.    The motor is supposed to just move at the tack speed until the boat's heading reaches the tack threshold before switching back to normal autopilot mode on the new tack.    This is simplified and obviously isn't the best strategy since tacking in different sea states need completely different settings for this.

With rudder feedback tacking could probably be improved, and it would be interesting to use rudder angle here.   The rudder can move to a known angle and wait until the tack threshold is reached.   Without rudder feedback this also may work better, by moving the rudder a constant amount and waiting.

Sean
Agree rudder angle would be ideal.
Need to think of what I really do while tacking but suspect the rudder angle is the same no matter what the conditions are.

I think specifying the time needed to get the rudder to the correct angle should work.
Should be  able to just put it back the other way the same amount of time as you approach the target heading.
Provably will take some tweaking to figure out when to bring rudder back and switch to regular control.

Have not used wind yet but only see using it  close hauled or when going close to down wind to avoid jibes.

How about a jibe preventer mode when on gps or compass to stop jibes?
Should have an alarm,