Ardupilot on OpenPilot Revolution - Page 5 - RC Groups
Thread Tools
Dec 16, 2016, 08:30 AM
Registered User
big plans yet deposited - they needs too much to rework. Version with Ardupilot's order can be uploaded if it needs
Sign up now
to remove ads between posts
Dec 16, 2016, 12:06 PM
Registered User
If you could upload a version with ardupilot order could be better.
Thanks

P.S.: I'am curious about your big plans
Dec 17, 2016, 02:17 AM
Registered User
Ok, uploaded. Plans was to give an ability to select motors layout via parameters but the parameters mechanism allows to work only with the child classes of class Copter, ie directly from HAL is not possible. So it is necessary to do a helper class.

PS. before reflash, don't forget to write all parameters to file - reflash clears all parameters so you will need to calibrate all again and again.
Last edited by night_ghost; Dec 17, 2016 at 05:24 AM.
Dec 19, 2016, 11:51 AM
Registered User
@night-ghost
The problem relative to my quad flipped upside down on little pitch / roll movement was not due to incorrect motor order, the second-last version of binary posted has the correct ardupilot motor numbering, I verified with Mission Planner that have a motor test function.
The problem is that the board results rotated of 180 to the firmware, I see this through the artificial horizon on mission planner and apmplanner2. If I set AHRS_ORIENTATION=4 (Yaw180) I see the right behaviour on the artificial horizon but the board is mounted with the arrow toward front and with iNav I don't have this problem.
I don't know if have explained well the problem but I think that is a software problem.
Let me know if you need more informations.
Dec 19, 2016, 02:23 PM
Registered User
Thanks for testing, I'll think it. Now my testing setup is almost done
Name: IMG_1910l.jpg
Views: 33
Size: 684.6 KB
Description:
RevoMini is on the other side of board, and full weight of brain is only 17 grams


UPD. Hmm, the sensor orientation is taken as-is from Ardurevo and was not changed. I of cause can change it but can you test ardurevo too?
Last edited by night_ghost; Dec 20, 2016 at 01:08 AM.
Dec 20, 2016, 06:53 AM
Registered User
Quote:
Originally Posted by night_ghost
UPD. Hmm, the sensor orientation is taken as-is from Ardurevo and was not changed. I of cause can change it but can you test ardurevo too?
I already did the test with ardurevo when I saw the strange behaviour on the artificial horizon: with ardurevo I have a correct behaviour with the default orientation, with your version I have to set AHRS_ORIENTATION=4 (Yaw180) to obtain correct behaviour.

Did you tested the two version on your revomini?
Dec 20, 2016, 10:59 AM
Registered User
Quote:
Originally Posted by anbello
Did you tested the two version on your revomini?
Only in MP artifical horizon, if I could fully test I would not be addressed here.

PS. I got it - ARHS orientation changed in ArduCopter since 3.1 So I'll correct it in minits
Dec 20, 2016, 11:55 AM
Registered User
Quote:
Originally Posted by night_ghost
PS. I got it - ARHS orientation changed in ArduCopter since 3.1 So I'll correct it in minits
Good

Pay attention to the fact that in the last binary version the motor order was incorrect while in the second-last version it was correct.
Dec 21, 2016, 12:05 AM
Registered User
Quote:
Originally Posted by anbello
Pay attention to the fact
Thanks, reverted, updated. But today my ST-Link died so I can't debug anything for some time
Dec 21, 2016, 07:18 AM
Registered User
Quote:
Originally Posted by night_ghost
Thanks, reverted, updated. But today my ST-Link died so I can't debug anything for some time
No ... what a bad luck
Dec 21, 2016, 07:22 AM
Registered User
Quote:
Originally Posted by night_ghost
Ok, uploaded. Plans was to give an ability to select motors layout via parameters but the parameters mechanism allows to work only with the child classes of class Copter, ie directly from HAL is not possible. So it is necessary to do a helper class.
This is really a motor mixer. I think this setting should go to the libraries/ap_motor folder.
Maybe we could just use existing FRAME parameter.
Dec 21, 2016, 11:41 AM
Registered User
Quote:
Originally Posted by Sh83
This is really a motor mixer.
do the procedure through the motor mixer can be, but ... it's like a bazooka on sparrows, too expensive and there is a way a lot easier. The more that the opportunity to have options for the HAL with the appointment of the conclusions would be useful in and of itself.
Dec 21, 2016, 11:45 AM
Registered User
Quote:
Originally Posted by anbello
No ... what a bad luck
Today I studied this issue and now is able to make a rough replacement from LeafLabs board, so I hope that I'll can debug it again in few days
Dec 24, 2016, 07:55 AM
Registered User
homemade ST-Link is working properly so show will go on!
Dec 24, 2016, 12:49 PM
Registered User
Good news)
By the way, port seems complete enouth. What should be done next? Sbus/dsm via softserial on ppm pin?


Thread Tools