2023 beta image

[Jean-Marc Douroux]

I have put a new image:

https://pypilot.org/images/

Thanks Sean
There is no longer any blocking of gain settings

[seandepagnier]

Please report any issues!

I am not adding features I would like to finalize a stable image.

[Jean-Marc Douroux]

Hello Sean,

I tested the 0.44 on the boat. I had no problems with calibrations and configuration with the tablet.
I also had no problem at sea adjusting the gains.

Thanks

[Jean-Marc Douroux]

Hello Sean
I have been sailing in recent days with version 0.44 and I still had a few rare blockages in the gain tuning fields with the tablet. Everything returns to normal by closing the browser and logging it again.

I can't set the automatic tack. I previously had an FF gain of 1.6 with version 0.24. I could turn quickly and precisely by pressing the +/-10 button 9 times. With the new tacking, the boat exceeds by 20 or 30° even if I reduce the FF gain a lot. But the boat then changes course too slowly by pressing a key 10

After changing from 0.24 to 0.4, some buttons pressed on 433MHz remote controls are sometimes not decoded even if pressed for a long enough time. They are received because the blue LED lights up on the hat. I also have some difficulty programming the +1 and -1 keys although less than with 0.42 and 0.43.
The hat is that of 2021. Should we change the arduino code to Pypilot 024 at 0.4?

I tried to update the arduino code using a PI 4 with the method described in the wiki. But there has always been a failure.

Fortunately, Pypilot still steers just as well with a lot of sea. In addition, it is incredible how little current it consumes with an old Autohelm tiller drive.

[seandepagnier]

Hello Sean
I have been sailing in recent days with version 0.44 and I still had a few rare blockages in the gain tuning fields with the tablet. Everything returns to normal by closing the browser and logging it again.

Is this with the new '<<' '>>' buttons for changing gains instead of the slider? It was supposed to be easier to tap than slide these.
Maybe refreshing the site fixes the problem rather than having to close the browser? Any hints on how to reproduce this?

I can't set the automatic tack.

Do you mean pressing the tack button? It should then ask for port or starboard tack or cancel. Did you find this?

I previously had an FF gain of 1.6 with version 0.24. I could turn quickly and precisely by pressing the +/-10 button 9 times. With the new tacking, the boat exceeds by 20 or 30° even if I reduce the FF gain a lot. But the boat then changes course too slowly by pressing a key 10

Hmm. Does the course change normally with FF gain of 1.6 with the new version? You are saying it overshoots by 20 to 30 degrees?

Is it at all possible to confirm the old version really has a different behavior here by trying in similar conditions? Is it possible that the wind strength or sea state differed? There really should not be any significant behavior change between the versions regarding this. So perhaps it would also overshoot by 20 or 30 degrees sometimes in the older version but sometimes not, and you remember it a different way? Can you confirm all the other gains are identical between the two versions?

After changing from 0.24 to 0.4, some buttons pressed on 433MHz remote controls are sometimes not decoded even if pressed for a long enough time. They are received because the blue LED lights up on the hat. I also have some difficulty programming the +1 and -1 keys although less than with 0.42 and 0.43.
The hat is that of 2021. Should we change the arduino code to Pypilot 024 at 0.4?

Which remote control are you using? Can you confirm that it programs well with 0.24 version? Are the +1 and -1 specifically more difficult?

There are only slight differences on the hat, mostly related to more precise manual control. You have not mentioned anything about manual control so I would suggest you do not need to update it.

I tried to update the arduino code using a PI 4 with the method described in the wiki. But there has always been a failure.

I am not sure what failure you are seeing, but you should be sure to stop the pypilot and pypilot_hat services before you can update the firmware.

Fortunately, Pypilot still steers just as well with a lot of sea. In addition, it is incredible how little current it consumes with an old Autohelm tiller drive.

It is always good to get positive feedback too!

[seandepagnier]

I found some problems

Hello Sean
I have been sailing in recent days with version 0.44 and I still had a few rare blockages in the gain tuning fields with the tablet. Everything returns to normal by closing the browser and logging it again.

I found an issue here and have corrected it.

I can't set the automatic tack.

Do you mean pressing the tack button?  It should then ask for port or starboard tack or cancel.  Did you find this?

I previously had an FF gain of 1.6 with version 0.24. I could turn quickly and precisely by pressing the +/-10 button 9 times. With the new tacking, the boat exceeds by 20 or 30° even if I reduce the FF gain a lot. But the boat then changes course too slowly by pressing a key 10

Hmm.    Does the course change normally with FF gain of 1.6 with the new version?    You are saying it overshoots by 20 to 30 degrees?

Is it at all possible to confirm the old version really has a different behavior here by trying in similar conditions?  Is it possible that the wind strength or sea state differed?    There really should not be any significant behavior change between the versions regarding this.  So perhaps it would also overshoot by 20 or 30 degrees sometimes in the older version but sometimes not, and you remember it a different way?    Can you confirm all the other gains are identical between the two versions?

After changing from 0.24 to 0.4, some buttons pressed on 433MHz remote controls are sometimes not decoded even if pressed for a long enough time. They are received because the blue LED lights up on the hat. I also have some difficulty programming the +1 and -1 keys although less than with 0.42 and 0.43.
The hat is that of 2021. Should we change the arduino code to Pypilot 024 at 0.4?

Which remote control are you using?    Can you confirm that it programs well with 0.24 version?    Are the +1 and -1 specifically more difficult? If it is only +1 and -1 functions I might know the reason. All of the other functions should not be affected.

There are only slight differences on the hat, mostly related to more precise manual control.  You have not mentioned anything about manual control so I would suggest you do not need to update it.

I tried to update the arduino code using a PI 4 with the method described in the wiki. But there has always been a failure.

I am not sure what failure you are seeing, but you should be sure to stop the pypilot and pypilot_hat services before you can update the firmware.

Fortunately, Pypilot still steers just as well with a lot of sea. In addition, it is incredible how little current it consumes with an old Autohelm tiller drive.

It is always good to get positive feedback too!

[Jean-Marc Douroux]

Hello Sean

1- Thank you for the correction to improve the gains tuning with the tablet

2 - Launching the tack works well. 

3 - I can't avoid a big overcourse with the same gains. I cannot compare automatic transfer with the old version 0.24 with which tack was not yet functional. I have overshoots of 20 to 30 degrees with FF gain between 1.6 and 0.5 with the new version 0.4 while the boat turned quickly and precisely before with FF gain of 1.6 and pressing the +10 or -10 button 9 times. I wonder if it's not because I have an Ap.tack-theshold of 79%. I just saw that on image 0.44, it is 20%. Unfortunately, I'm no longer on the boat to check

4 - I have been using waterproof 4-button 433MHz EV1527 remote controls for more than two years. I never had a problem before using version 0.4. I wonder if it wasn't +1/-1 and +10/-10. The computer display is not visible from the cockpit. I had to juggle with the tablet between tuning and control. It's weird because it works 10 times and then sometimes it doesn't.
However, I saw that you had already improved this problem a lot between 0.41 and 0.44. With 0.41, I had a lot of trouble programming the keys. I think there is still a slight problem with the +1, -1, +10 and -10 keys 

 5 - Ok. I don't use manual control because the radio remote controls worked well. I have a remote "engage" "disengage" on the tiller cylinder. I never needed to consult the hat screen while sailing.

6 - Bravo for what you did. I have much more confidence in your vision and Pypilot than in the pilots I have sailed with before.

[seandepagnier]

3) The tack settings only affect tacking. If you were manually tacking by pressing +10 9 times there would be no effect of the tack settings. I really suspect there is no difference in the versions but a difference in the sea conditions, wind, sail trim, compass calibration or something else. There are a lot of factors. I hope in the future to have smarter pilots that automatically tune themselves to provide more accurate control.

4) There was a small change I fixed for 0.45 with specifically +1 and -1 I had some experimental code to make manual control faster for these and that may have been why. In 0.44 it should work fine to program them, just not display the action in the web when doing so.

5) Ok. Manual control may not be useful to all users.

Thanks again for the feedback. I will try to place a new image tomorrow with these improvements.

[seandepagnier]

I have put a new beta image that should solve most of the problems mentioned. Please test!

https://pypilot.org/images/tinypilot_bet..._24.img.xz

[MarineW29]

I am still attempting a final release for pi zero boards  (pi core 9)  after which I will shortly be moving to pi core 13.

Hi Sean,
When do you plan to have a stable image of tinypilot based on piCore 13.1.0 or above, which can support Raspberry Pi Zero 2 W ?
I bought a  Raspberry Pi Zero W, but they sent me the version 2, and it took me a while to understand why it didn't boot at all (same problem as described here).
Thanks