if some one have an idea to get gps to work on im ready for testing

Check your PM,Thanks

Servo for my camera stabilization are twitching, making small and anoying movement on video.
They were also connected directely to radio receiver for testing purpose, and no twitching appear, it seems that problem is coming from parasite either generated from code (I'm using v2.3) or maybe from the board (strange).
Does anybody had the same problem ? How to correct it ?
Thanks

Originally Posted by HS6MTU

Check your PM,Thanks

look your pm

I put megapirates 2.4.1 on my black vortex and I noticed that throughout the flight a random motor would suddenly twitch. It has never been enough to cause a crash but I was curious if this is a bug that others have recognized or is something else possibly causing my problem? It is as if the pwm output from the controller to the motors occasionally sends an incorrect value, causing the esc to send a single rapid pulse of throttle to that motor.

sg1anubis: what do you mean by "twitch"? There's a random yaw issue which could look like twitch... Depends how many motors you have, it will look different on quad or octo...

It is as if the throttle of a single motor will try to jump from 50% to 100% or 0% for only a fraction of a second and then back to 50%. It is very brief and causes the aircraft to pitch maybe 20 degrees for only a moment. It has never been enough to cause a crash and it seems to have only begun happening after updating to 2.4.1.

I'm flying a fairly heavy quad with a weight of about 1.55kg with the battery and gopro mounted. I recently changed to some 11x4.7 props. You can see my setup a few posts back:
https://www.rcgroups.com/forums/show...postcount=1482

On an unrelated side note, I recall the first time I tried loiter mode with 2.0.49 it violently pitched and nearly crashed after triggering loiter. I had a solid GPS fix but had never adjusted the pids. What behavior should I expect from loiter mode with unadjusted pids? What is the proper way to go about tuning the navigation pid settings?

EDIT: I just reverted back to 2.0.49 and after 40 minutes of flying the twitch was gone. The ESCs that I'm using are 30A blue series flashed with SimonK's firmware so I'm guessing there is a bug in the code that is intermittently sending bad signals to the speed controllers. The refresh rate on my speed controllers is higher than standard escs so it makes sense that normal escs may simply reject the bad signal whereas mine pass the bad signal on to the motors.

With high refresh escs I have found that if you make rapid changes in throttle, for example 0-100% nearly instantly, then the motor locks up or freezes for a fraction of a second. When 2.4.1 started showing those behaviors I thought that a bad signal may be the problem. I imagine that the pulse widths of the signal may look something like this while hovering:

1520,1523,1521,1522,1520,1520,1519,*1948*,1520,152 2,1523

sg1anubis:

1) twitch - do you have CH7_RTL or something similiar set? maybe even connected CH8 to the controller? (not sure if they left passtrhrough in there but it should _never_ be connected if they did) I had something similiar happen to me - i didn't calibrate _all_ the channels and when the signal jumped a little it triggered whatever was set there (the same would go for CH5 - if you have a border value for a mode). Another possibility is bullet connector going bad, ESC going crazy... I don't think it's controller related really
2) loiter - it shouldn't go crazy right away, it should jump ramp up some speed, overshoot, go back, overshoot... If it works well in stabilize it should not behave that crazily in loiter (certainly not in 2.0.49) - to reiterate - are you sure it was really loiter and not something like ACRO? (is your CH5 switch set up right?)
3) ah - simonk's ESC firmware... it should work fine in the latest "stable" versions, but I never used it, I just follow the thread somewhat - I know there were serious issues in pre-December versions of the firmware, and I know motors locked up when throttle was upped to fast - this is NOT good for multicopter either way because even if you don't see it because there is some momentum involved, the controller can send large changes fast and cause something like this. Original ESC firmware does some filtering and smoothes sudden changes, simonk's does no (or at least not that much) filtering. I'd rather follow up with him and see what he thinks of it. Can be a glitch, synchronization error.

Btw older versions of AC2 had a thing called "instant pwm" which I think is aimed at ESC firmwares able to react faster to changes, so every new calulation is sent to them. Default rate was lower (100Hz? something like that). 2.4.1 has 490Hz set for "not instant" pwm (I don't get the difference, works for me and didn't investigate), maybe that's part of the problem... I think you can change the rate to any value, so try 200Hz and see if the twitching goes away.

Ya, one of the first things that came to mind was the issue that people were having with earlier simonK firmwares, that could be contributing to the problem; however, it would mean that the signals coming from the controller weren't just changing rapidly but greatly fluctuating in range, this shouldn't happen during a stable hover. With multiwii I tested with the highest possible motor resolution and refresh rate without any twitching, it makes sense that if the refresh rate of 2.4.1 is higher it may cause the twitching because I have never had any lock up issues with any other software or version of megapirates. Ultimately, it isn't a big deal, it rarely happens and has never caused a major malfunction, I just thought it may be worth pointing out. I'll have to test out loiter this weekend when I have a more suitable location to fly. Are you using default pids for navigation and loiter?

Hi all.
Finally I received my new Board. But now I have a problem.
I hope you can help me. I installed on Black Vortex board the mega version 2.049. My quad not fly stable.
Did the P value of 0.3 and it's a bit better. I value and IMAX values ​​are still default.
I've replaced landing gear. Engines and propellers are balanced. The gas paths from the controllers are not programmed. Is this the problem?

Quadro problem (0 min 24 sec)

Sorry for my English.

Hope you can help me.

Regards
Georg

Originally Posted by zviratko

Btw older versions of AC2 had a thing called "instant pwm" which I think is aimed at ESC firmwares able to react faster to changes, so every new calulation is sent to them. Default rate was lower (100Hz? something like that). 2.4.1 has 490Hz set for "not instant" pwm (I don't get the difference, works for me and didn't investigate), maybe that's part of the problem... I think you can change the rate to any value, so try 200Hz and see if the twitching goes away.

InstantPWM used in MPNG until now. Today I will release 2.4.1 R2 where FastPWM (400Hz) reverted and can be enabled.
InstantPWM works at 150Hz+ (Depends on main cycle timings).

Hi Sir Alex
What other changes have the r2 v

Originally Posted by Sir Alex

InstantPWM used in MPNG until now. Today I will release 2.4.1 R2 where FastPWM (400Hz) reverted and can be enabled.
InstantPWM works at 150Hz+ (Depends on main cycle timings).

Sir Alex: ah, so the "as soon as calculated" is actually slower? I was wondering if their cycle is really so fast (compared to MW which has 2500 microsecs typical)

Originally Posted by sakiskoz

Hi Sir Alex
What other changes have the r2 v

http://code.google.com/p/ardupilot-mega/source/list

depends at what point Sir Alex ports the code, but usually it's the most recent version at the time of the release


Sir Alex: thank you!!

Originally Posted by sakiskoz

Hi Sir Alex
What other changes have the r2 v

For BV boards:
* Fixed small bug in config.h. Now you can just set PIRATES_SENSOR_BOARD as BV, GPS and sensors will be selected automatically (due bug it's not work in previous versions)

For all boards:
* Restored FastPWM. Default 400Hz, and can be changed in APM Planner via RC_SPEED value. But InstantPWM still selected by default (in APM_Config.h)
* GPS initialization fix. Many time I'm trying to fix my MTK3329 GPS, with binary protocol GPS_MTK16. Now it's work fine (at least for me).
* Apply InstantPWM Syberian's fix. It must remove jitter on PWM channels.


MPNG 2.4.1 R2 = AC revision