Servers by jDrones

Sailboat Support

This rudder issue must be sorted out first I think.

Today I did total recalibrations of accelerometers and compass as well as on the tx sticks and the radio input channels. I did it after having a good gps 3d fix outdoors.
Trying out the rudder in different modes I discovered something that can explain a lot of the strange behaviours I have experienced before all along, both when sailing the soft sail and with the wing sail.
Have a look at this

  1. This is Acro mode when unarmed - rudder responds as expected with tx stick movements.
  1. This is when just arming the boat in Acro mode - weird rudder responses to tx sticks.

The only difference between the two videos is that its unarmed respectively armed.

I have tried several times after rebooting the boat and the pc and the tx.
Result is the same. It does not go away.
Its impossible to sail with this in Acro mode.
Unless I leave the boat unarmed, but that will likely have some other effects on logging and RTH, missions etc etc

@iampete you have any idea what can cause this?
I have logs and setting file if that will help…

p.s.
No warnings from GPS or mags or bad sensors during this time when weird rudders, all looks stable and fine.

Hum, hard to say if its a real issue, in acro mode you control direction its trying to sail in but because its not getting any response on the stand it does unexpected stuff. Possibly on the water it would be OK. I dont think there should be any difference between armed and disarmed in acro, maybe the steering interrogator term of the PID is zeroed when disarmed.

Thanks Pete
I think I know by now that it will not be fine in the water since so many unexpected things has happen since I started months ago.
did you ever try that on your boat?
I bet it will not behave like in the video.

what setting is that? where can I check that?

Today I was at the lake to find out if it possibly could be okay when its getting response on its stand.
Unfortunately it was too risky to go out since when no gust I had some 10-12 knot of wind but in the gusts it was over 20 knots. I did not want to risk the boat doing that.

@iampete you have any chance to test on your boat if it behaves the same way as in my previous two short videos on armed respectively disarmed in Acro mode?
It would be very helpful. If your boat don’t react like that then I know its something in my hardware or sw configuration.

I don’t get much time on the water, but testing on the stand I can do, I will have a look this eve

testing on the stand is all I need
Thanks a lot Pete!

This is my sailboat with the rudder reversed in acro mode, this is because it had no gps lock so though it was sailing backwards. In maunal mode it is correct, and on the water with GPS lock and moving its fine.

The issue is on the stand firstly the boat is going very slowly so Rover used full rudder inputs. It may be more intuitive to think that it used less rudder as it goes faster, so we have to use max at zero speed.

The second issue being on the stand is that the GPS position wanders about so sometimes it thinks its going forwards, sometimes backwards. As soon as we start going backwards rover reveres the steering output so as to still turn in the same direction. This can lead to the rudder flip flopping between full left and full right.

Once your moving this GPS jumping about results in small changes of velocity rather than direction. ie 0± 0.1m/s is a change of direction, where as 1±0.1 is not. Also once moving the EKF sort of has more to work with so it gets a better estimate.

Thanks a lot Pete, so great to have a logical explanation.

May I suggest, since sometimes out on water there might often be situations similar to be in a stand on the dock side, that you create a logical switch like REVERESE_RUDDER_IF_NOT_GOING_FORWARD = 1 or 0
Would be so great to be able to remove this uncertainty from the sailing.

It would be nice to have a proper fix for this, basically because its not moving there is not point in steering. The issue is that its quite hard to tell when stopped becomes moving very slowly. We could as you say just had a parameter to stop it reversing the steering.

my only 10 cent opinion is that its problematic to have a manual that say in Acro the pilot controls the rudder and the fc controls the sheeting. Then it is not true at all.
The fc is in fact tampering with the rudder.

in acro the pilot controls the turn rate rather than the steering directly, auto sheet and manual rudder might be fun tho, i think we would have to have a new mode. I think there might be a parameter that allows auto steering with manual sheet.

maybe its easier to make a new mode than to change the current Acro mode but what I miss is not a manual rudder but a ‘simple and clear’ acro mode that just keeps the course using the stick rate and not change the rate depending on the ground speed or direction of the boat. I think that will be a very enjoyable and relaxing sail mode.
Anyway, now I understand better what is going on and what to expect.

Last weekend had better wind, not so shifty and just a bit stronger.
During this ~40 min video I had no gps or magnetometer warnings and the wind instrument seem very reliable even in weak winds.

(+) Acro mode seems to work as expected with the Wing. When I use rudder to tack it does flip the wing tail over in the correct way. Consistently.
Most of the time it succeeds to tack, if boat speed was too low then it could get stuck half way and fell back to original tack. But the wing tail was correct so just had to gain more boat speed and retry.

I think Pete mentioned before that the throttle stick on the rc should be in top position (sheeted out max) for the boat not to try sailing backwards. In my set up it works correctly having the rc throttle stick full down (max sheeted in). When I changed it around, then it worked backwards.

I think this is an achievement, this part of the code works - well done!

(-) I could not get Auto mode to work.
I am new to waypoints with Ardupilot so maybe I set it up wrong. Idea was to have two waypoints and the mission set to have the boat sail from 1 to 2 at beam reach and then just repeat that 100 times. Was planning to observe how it would perform and maybe make some trims.
This was uploaded to the flight controller but when I switch to Auto mode the boat just make circles. I try this 4 times in the video. The first time it seems it aims for the Home position, rather than wp 1, but I am not sure. Maybe it should?

(-) To get rudder steering in correct directions in Acro when Armed, I had to set the rudder direction so it was working in the opposite direction when in Manual unarmed mode. So when switching to manual mode out on the water, then the steering would be backwards, bit awkward. I feel really confused here.

hum, great that the wing sail is working!

not sure what is going on with auto, you can see its trying to go to WP1 because it puts the orange to current way-point line on it. Typicaly if acro works then auto works, maybe it was trying to reverse the whole way. Because you were having to put throttle down not up.

Sounds like some of your issues could be a mix up of reversed inputs and outputs, first step is to check all the rc input green bars are moving the same way as the sticks (except pitch but that doesn’t matter for us). Then you should change the rudder reversing such that its correct in manual mode, this should then be correct also in acro. Then in acro with stick fully up you should reverse (or not) the wing sail elevator such that it sails forwards.

thank you. I got it.

A little teaser of things to come…

2 Likes

that’s a trimaran with foils right?
seems to have speed resources :slight_smile:
foiling mini 40?

1 Like

Just working on the bench at home to get prepared for sailing now,

Completed fine.

completed fine.

But,
is it really okay that if I arm it, then in manual mode, the rudder will turn the opposite way?
Is that something just to ignore?
This is not really cool and very counter intuitive. Why? Don’t try to answer that question :slight_smile: - but I feel it should be turning the same way when its armed as well - why not?

yep!

this is not OK. Must be something odd going on. Armed in manual is backwards but disarmed it is correct? Might be worth updating mission planner, there was a bug that it would not display the reverse input channels correctly. The way to be sure is to turn off all the rc reversing params ( RCx_REVERSED) and do any reversing required on the transmitter. Or you could double check with QGC.

Where did you get the mini 40? I know of a guy in New Zealand, I think, who has done all sorts of work to be able to adjust the foils.

Servers by jDrones