Thread Tools
Mar 10, 2014, 10:14 AM
Registered User
R_Lefebvre's Avatar
Thread OP
Quote:
Originally Posted by charosenz
Ben,

If I am reading your logs right, it looks like your Rate P and Rate I is something like .798520 on both. That is really really high. Most run around .10 to .14. Can you double check that? The Rate D at .004 seems normal. Of course these are very different for every machine so I am just ball parking the numbers......
Good catch! Yes, that Rate P is INSANELY too high. Some even run down to 0.05.
Sign up now
to remove ads between posts
Mar 10, 2014, 10:16 AM
Registered User
R_Lefebvre's Avatar
Thread OP
Quote:
Originally Posted by Xoltri
Ben, have you checked for motor sync issues? I'm battling that with my Pixhawk resulting in a few flip over crashes. Finally was able to recreate the issue by strapping the quad down, arming it, then playing with the throttle. If I go from 0 to a high throttle setting using my throttle hold switch I can get the motor to stutter and fail to provide power. So until I get this issue resolved there is no point in flying as it will just crash again.

Might be worth it for you to do this test to see if it is your issue as well. I'm using the SimonK ESC's. Been documenting my troubleshooting here and here.
Yeah, the SimonK thing is a bit of a problem. There are a few solutions. You can play with the settings in SimonK to make them work better. Or use standard firmware in the ESC's.

We talked about putting a damper on the motor outputs, but then all we're doing is eliminating the benefit of running SimonK in the first place.
Mar 10, 2014, 10:36 AM
Travelling into the future
Xoltri's Avatar
Is there any difference between how the APM communicated with the ESC's vs the Pixhawk? I didn't seem to have this issue with my APM, it's only showed itself since I moved to Pixhawk.
Mar 10, 2014, 10:39 AM
Registered User
R_Lefebvre's Avatar
Thread OP
No, I don't think so. You have the exact same PID's, etc?

The only thing I can think of, is that I believe there is a difference filter running on the IMU data. The gyros should be "sharper" while also being more resistant to noise. I don't know a lot about it other than that's what Leonard says. But if the data is "sharper" then the output might also be sharper coming from the PID. You could detune the PID's a bit to try and soften it a bit.
Mar 10, 2014, 11:07 AM
Registered User
Clippedwing's Avatar
A few things to test tomorrow, this is a great start! I don't remember making the P gains like that!

Now that you mention it, i will look into the motors aswell! Strapped down with props on?

Did you watch the video i posted on my instergram? It flew perfectly with the same motors/ESC combo or can this just happen?

Thanks again!
Mar 10, 2014, 11:24 AM
It's Just a Hobby!
Sudsy's Avatar
Quote:
Originally Posted by kallend
I have a spare APM2.6 looking for an application (my other 2 are used in planes) so I thought I'd try a multicopter for a spring project.

I already have 2 NAZA equipped multis on DJI frames, and I thought I'd try a different frame for the APM.

I figured the first APM attempt should be something inexpensive with a proven track record with APM, since the NAZA know-how doesn't really cross over.

Any recommendations (frame/ESC/motors/props) that are known to work well?
Have a look at the Hobby King Talon V2 frame. I had an early version that I flew with a KK2 board for quite a while. I crashed it and ordered another frame kit thinking it was the cheaper way out instead of ordering individual parts. Glad I did. The new kit had some very nice improvements including Hex Driver screws instead of Phillips screws. This is a major improvement in that it allows the arms and plates to be much more snuggly fixed. I use NTM propdrive 28-30 750watt motors and 30 amp simon K esc's. I have another frame and if I put one of the APMs I have on it I am going to use a set of NTM 28-36 850?watt motors and regular 30 amp esc's. Props are 8x4.7 plastic on the first one here and I'll would put 11x5 props on the next one.

Sudsy
Mar 10, 2014, 11:31 AM
Travelling into the future
Xoltri's Avatar
Quote:
Originally Posted by BenLimeClouds
A few things to test tomorrow, this is a great start! I don't remember making the P gains like that!

Now that you mention it, i will look into the motors aswell! Strapped down with props on?

Did you watch the video i posted on my instergram? It flew perfectly with the same motors/ESC combo or can this just happen?

Thanks again!
Yeah you need the props on to recreate the issue. To narrow it down I just have one prop on and only that motor connected to the flight controller. I built a board that I can strap the quad down to so it doesn't move around, and I can stay far enough away from it to be safe.
Mar 10, 2014, 11:32 AM
Travelling into the future
Xoltri's Avatar
Quote:
Originally Posted by R_Lefebvre
No, I don't think so. You have the exact same PID's, etc?

The only thing I can think of, is that I believe there is a difference filter running on the IMU data. The gyros should be "sharper" while also being more resistant to noise. I don't know a lot about it other than that's what Leonard says. But if the data is "sharper" then the output might also be sharper coming from the PID. You could detune the PID's a bit to try and soften it a bit.
Thanks for the advice. I'm just using the default PID's. I tried to run autotune but it resulted in a near crash due to the sync issue so I need to get that fixed before trying anything else.
Mar 10, 2014, 12:46 PM
Registered User
Clippedwing's Avatar
Quad Problem APM (0 min 12 sec)


I forgot my friend recorded some of my flights! This will help loads!

This is a flight, default settings using channel 6 to tune roll/pitch on the max of 0.2


I notice it lifts off fine, then it just wobbles like that, I tried to tune it out I just could not!
Mar 10, 2014, 06:16 PM
AMA 697691 / FAA 13675
Hi Robert,

I'm pretty happy with my APM 3.1.0 performance so I am hesitant to just upload 3.1.2 without first asking...what are the changes? Should we bother with 3.1.2 if using an ArduFlyer 2.5.2 board?

Thanks.
Latest blog entry: Greg's Web Hangar
Mar 11, 2014, 12:11 AM
Registered User
charosenz's Avatar
Quote:
Originally Posted by R_Lefebvre
So I talked to Randy yesterday. The listed range of +/- 300 is a mistake. It should probably be +/-1.0, but typically these fall in the range of +/- 0.5. He said he'll fix it for 3.2.

So, good question.
I appreciate the clarification, but does this mean is should be between -1.0 and +1.0...Or?
Mar 11, 2014, 12:37 AM
Travelling into the future
Xoltri's Avatar
Well I spent another 2 hours on this tonight and I believe the issue is not with my speed controllers but with the Pixhawk, or with how the Pixhawk communicates with the speed controllers. See the videos below. When I have the ESC connected direct to a receiver it works fine. When I have it connected through my old APM it works fine too. But when I connect it via the Pixhawk and do the same test I get motor sync issues.

Pixhawk simonk sync issue (2 min 5 sec)

Follow up to pixhawk sync issues (1 min 15 sec)
Mar 11, 2014, 05:53 AM
Registered User
Clippedwing's Avatar
Quote:
Originally Posted by Xoltri
Well I spent another 2 hours on this tonight and I believe the issue is not with my speed controllers but with the Pixhawk, or with how the Pixhawk communicates with the speed controllers. See the videos below. When I have the ESC connected direct to a receiver it works fine. When I have it connected through my old APM it works fine too. But when I connect it via the Pixhawk and do the same test I get motor sync issues.

https://www.youtube.com/watch?v=i7_gVoY8q8M
https://www.youtube.com/watch?v=VTK81IANyUE

Wow, that sounded harsh via the PixHawk! It almost sounds like it was running in reverse!

I hope you get your problems sorted out soon!
Mar 11, 2014, 08:53 AM
Registered User
R_Lefebvre's Avatar
Thread OP
Quote:
Originally Posted by Xoltri
Well I spent another 2 hours on this tonight and I believe the issue is not with my speed controllers but with the Pixhawk, or with how the Pixhawk communicates with the speed controllers. See the videos below. When I have the ESC connected direct to a receiver it works fine. When I have it connected through my old APM it works fine too. But when I connect it via the Pixhawk and do the same test I get motor sync issues.

https://www.youtube.com/watch?v=i7_gVoY8q8M
https://www.youtube.com/watch?v=VTK81IANyUE
That's really interesting. Thanks for researching this. I'm bringing it back to the other developers.

Couple questions, what is the setup? You have throttle on the stick, and you're using a throttle kill switch to alternate between low and high? So the change is instantaneous?

Now, you're sure you have the same throttle output in all cases? It would be best if you could confirm the actual PWM output to the ESC to be sure.

Which ESC and motor are you using? Just so we have all the data.
Mar 11, 2014, 08:54 AM
Registered User
R_Lefebvre's Avatar
Thread OP
Quote:
Originally Posted by charosenz
I appreciate the clarification, but does this mean is should be between -1.0 and +1.0...Or?
I don't think there's a hard and fast rule. But if you've got more than +/-1, we might want to have a look at that.


Quick Reply
Message:

Thread Tools