Just downloaded 2.0.8 from the planner. The winds here in the UK are making a real nuisance of themselves at the moment but despite the weather I was able to do a brief test and it flew okay in Stable mode. However, I see that the yaw 'unwinding'* issue is still there. Several people have mentioned this and yet it has remained a small but noticeable problem for some time. Would it help if I made a short video of the problem so that the project team can see what is happening or is it a known and difficult issue to resolve?
* The yaw unwinding issue as far as I'm concerned is where you provide a yaw command, the copter then rotates but when you release the rudder stick the copter rotates back in the direction it had come from around 10 or 20 degrees.
I would like to see a video Nigel.
Interested to hear that you've tested 2.0.8 ok, I might be tempted to upgrade from 2.0.5 now, I was holding off because of some of the scary bug reports in the last week.
Yes the wind is a nuisance isn't it, I'm in Southampton and it hasn't really let up all week.
Do you know if the 2.0.8 version now has motor LEDs turned on by default? Jason added support in 2.0.6 but turned off by default, and I saw a report that when someone tried to turn them on they got a compile error.
We just went into the garden where the wind is even higher than it was this morning (so apologies for the poor flying) but we have published a video of the effect I was speaking about. It's rather hard to see because the copter is bouncing around so much but have a look at this video between seconds 10 and 14. You should just be able to that I apply left rudder, the copter yaws to the left, I release the stick and the copter yaws back to the right again. We rigged up a big pointer (which unfortunately worked in the opposite direction to the way we were moving the stick!) so you can see what the rudder channel is doing...
Interestingly, on a right yaw the copter just seems to keep going for a further 10 or 20 degrees - the opposite effect seen on the left turn.
I tried 2.0.7 yesterday, and yaw, although it responded quite crisply, with mag disbled, kept it's orientation; asap as I enabled the mag, yaw started to have it's own will (sort of speaking).
Have you tried it with mag disabled?
Ok yea i see what your are talking about. I noticed this to with the PiratesNG, I haven't been able to fly for the last 4 wks, SO i haven't been able to do any flight test with new AC2. I have noticed that the compass is becoming more and more accurate with every new AC2 firmware update though. Like Dani When I was using the Pirates code if i diabled the compass the yaw control seemed to function alto crisper. Has changing the YAW PID's mad any difference or do you think it has something to do with the code?
I'm going to implement an alternative (not as fancy) yaw control tonight. I'll put a compiler option in the APM_Config.h and you can try alternate ways. That'll help me sort this thing out for good.
I also added better Yaw logging today to help as well. If you have log issues, please erase your logs. That seems to help.
The problem I'm having with yaw is when I do for example a fast move to the right and level again, it starts to yaw left for about a 100° without stick input. If this happens a little far you loose orientation immediatly. If it flies stationary it doesn't do it. Strange. This happens with mag on and off.
Went back to the pirate code and everything flies just fine