Hi All, I am in Cape Town.
Built a 650 quad with pixhawk FC
Flying in stablz etc.
I cannot really eval whether the kit is properly 'tuned' since I have very little experience.
Is there someone who can help me 'hands-on' to eval my rig?
Thank you.
You need to be a member of diydrones to add comments!
Replies
contact me George 079 five two four 38one six can help youlive in Kraaifontein
Hi George,
Thank you, I will contact you to touch sides re my quest :)
Regards,
Gerhard
Not so Gerhard...
I'm in Cape Town and I'd very much welcome a chat!
I've reverted back to my APM 2.6 on my 550 cannibalized frame, as I've had issues with my RcTimer Pixhawk clone, which has been sent back for evaluation.
Drop me a line:
david@designersmedium.co.za
Quadcopter.jpg
Hi David,
Great to hear from someone on DIY from Cape Town.
Thank you for your contact-details,
I would like to touch base with you. :)
Regards,
Gerhard
Hi Gerhard,
From what I've read it is definitely worth it to AutoTune your multirotor and it works well in most cases and I'm sure it can work for your system as well. But I would be careful doing it the first time. That's why I think it's a good idea to test it on a test bench first to minimize risks.
I can only comment on my experience with autotune. I don't want to go in too much details on this thread - there is a long discussion in the link I gave you and I will have more information once I have my quad repaired, but in my case I have found a case where autotune didn't play well with my initial manually tuned PID settings and I think I understand what caused the problem.The big lesson I learned is that getting your quad to fly well manually improves your odds of getting autotune to work, but it doesn't guarantee it. This extract from my other discussion will show why:
The key lies in this graph:
This graph shows the actual roll angle (green) and the desired roll angle (red).
The left shows what happened in AltHold mode when I flicked the roll control stick to its limits and let it return to centre again.
The right shows what happened when autotune directed it to roll and then let it return to horizontal again.
The actual roll (gren) is fairly similar in both cases. Both resulted in a fairly big (20 degree?) roll that happened rappidly.
But look at the desired angle (red).
In AltHold mode, the desired roll change because I directed the quad to roll. The actual roll follows closely. Because the actual roll reacts to the desired roll, both actual and desired are always close to each other (small P error) and the slope of both are fairly close to each other (small D error).
In Autotune, the desired roll stays at 0 while the actual roll changes a lot. Because of this, actual an desired gets far apart (large P error) and there are big differences between the 2 slopes (large D error).
To visualize this, I graphed the P and D errors for both cases in this graph:
From this it can be seen that the P (red) and D (green) errors are a fair bit bigger for Autotuning (right side). But what makes matters worse is that in AltHold P and D often work against each other. If one is positive, the other is negative. So when the sum of P an D terms are calculated they often cancel each other out to some extend. In Autotune, both P and D are often reinforcing each other. This graph shows the sum of the P and D terms for both cases:
This graph shows the effect of the combined P and D terms. This reflects the size of the correction term that will be send to the motors. The maximum error value in Autotune is about 6 times the AltHold value. That is why (in my case) the same PID values with the same size disturbance flies well manually in AltHold, but starts to oscillate in Autotune.
Simply stated: My quad used fairly big PID values (relative to it's weight and motor strength). This means that it would rapidly correct for any disturbance it encounters. Because it always follows the desired position closely, the error always remains small and a small error multiplied by big PID constants result in reasonable correction values for the motor to follow. But the moment it encounters a big error (like in the autotune case, or if for some reason it would find itself out of position during flight) the error term would become big. Multiplying a big error with big PID values results in a big correction term causing it to overshoot an oscillate.
I think this should mainly be a problem for quads that can change it's rotation rapidly. Lightweight, short arms, strong motors. I think for quads with more rotational inertia the difference will be smaller.
That's why next time I manually tune my quad, I will start on a test bench. And I will be introducing large errors instead of just flicking the RC controls around and see if it can correct from a large error state. Once it passes that test, I can autotune it on the bench without any risk of crashing again.
Plenty of people here in South Africa using Pixhawks Gerhard, its all in the manual as Jack has pointed out. Yours is not a particularly big build. You might need to autotune your attitude a bit if you want help ;-) I'm in KZN otherwise I would help you tune it. Autotune is like magic.
Hi Gary,
Thank you for your interesting comments re my question.
I stand corrected, there is life on tthe SA DIY branch! ;)
Gerhard
Cool, autotune is not hard just make sure you are in a wide open area on a calm day.
If in any doubt stop and try again don't let it be the boss!
Hi Gerhard,
I'm in Adelaide (Australia), but originally from South Africa - so cannot help with hands-on, but I'm also in the process of tuning a home built quad using a pixhawk FC, so I can share some of my experiences.
You can check the arducopter autotuning wiki here:
http://copter.ardupilot.com/wiki/autotune/
Make sure that you check that the frame is sound before you attempt to fly. Specifically:
1. Check that the arms are rigid.
2. Check that vibration levels are low. You can use Mission Control's real time monitor to check this.
Before you do auto tuning, you must be able to fly the quad in stabilization mode. So you have to do some PID tuning by hand first. I would rig the quad in a 'test bench' that would allow it to only rotate in one axis (roll for example) and then tune it by hand first. There are some guides available for that. I found this one useful:
http://diydrones.com/forum/topics/arducopter-tuning-guide
Once you get some decent values on the test bench with manual tuning I would even attempt a autotune on the test bench to get some starting values before you take it out for free flight, then check that you can recover in stabilization mode in free flight and do another autotune to fine tune the free flight values.
The test bench approach may sound a bit over the top, but I wish I have done that. I got the quad to fly decent using manually tuned PID settings, but crashed as soon as the autotune process started. You can read my thread here:
http://diydrones.com/group/arducopterusergroup/forum/topics/autotun...
-
1
-
2
of 2 Next