AQGC and Calibration with DIMU very hard !!!

[plexus]

First flight with dimu and param of Jussih (thanks to him) very very good.
AQGC 1.3,
Firm R380 or R372
Dimu 1.1

BUT I had had a lot of difficulties to do my parameters:

1)
Abort on sim3 I read to load sim.exe and cal.exe, ok done.
Write an command, OK done and done each time I launch QGC 1.3.0
But result 'NAN'
____Abort before the end of the loop
Sim 5 a lot of "abort"

and now I cannot load my params. params in the AQ6 crash of QGC !!!! I dont see why.
With an other file no problem.

2)
In QGC 1.3.0 I dont see anywhere to choice CAN or PWM in connection .
I can see the icon only when the qgc is not connected (see attached file), after connection they disappear. I dont know if I am in PWM or CAN.

no connection icons are present

With QGC connected no icons

Many thanks for the explanations.
Always happy with AQ !!!

[Max]

Since the MOT_CAN param was renamed (MOT_CANL), QGC doesn't know how to set the CAN option, so it hides it. Updated QGC is on its way (code already in GitHub repo if u can build your own), but in the meantime you'd need to set MOT_CANL manually. There's another thread about that somewhere...

-Max

[Jdmagoo]

[plexus]

Hi and thanks for the replies,
Menno said:
Re: Can't find MOT_CAN
Postby joebar.rc » Fri Apr 25, 2014 8:44 am
If you have mot_can than you don't use the latest svn release. That's fine and you don't have to worry about mot_canl &h. Some time ago support for uart over can for the future pdb was implemented and that required the split in can_l & h params.
so... When you use a new release you will see them...

My config hexa+:

MOT_can1.jpg (29.24 KiB) Viewed 16142 times

Well, If I well understood, the use of the release of QGC is not necessary, using all motors with CAN.
In the attached file you see MOT_canL=63 (Hexa+), is all motors are automatically in CAN ?

Thanks
Alain

[joebar.rc]

[plexus]

Thanks menno,

If I had an quad and I want 2 motors in Can (motors 1 and 2 for example) and 2 in PWM (motors 3 and 4), I write CANL = 3
and automatically motors 1and 2 are in CAN and 3 and 4 are in PWM ?

8421
I I I I

Alain

[Jdmagoo]

[plexus]

Many thanks,
But now I have a new problem Sim2 abord after a short time !!!!
Tested with qgc1.4
grrrrr....

[Max]

Damn, you guys are quick... just posted that version last night for some people to test it out. There could well be some bugs and I haven't published any release notes for it yet... So YMMV The included calibration utils have been updated as well. Please do report any problems you find.

Thanks,
-Max

[plexus]

Hi Max
Thanks for the 1.4, I like the "dring,dring" for low batt.
i loaded the paramsj of Jussih and the new QGC seems to run well. Some flight marvellous.
But i dont understand why I cannot run the calculation of my dynamic aq log. On step 2 and 3 the loop abort, on step 4 some ´NAŃ´????
The dynamic log seems to be good but the result of the MAE is not correct: as 12345678.+e10.7. As if the dimu create a bad value.
am I alone to have these difficulties ?
the file, if anyone could have a look :

https://www.dropbox.com/s/jqdrn00nexyog ... re-AQL.zip

I do as in the wiki:
http://autoquad.org/wiki/wiki/autoquad- ... -pictures/

The old calculation without dimu had runned very well but with Dimu the values are not good.

After a night of brain storming:

My Dance

After redo several times the dance to be as close as possible than curves in the wiki, I redo the calculation.
Using the 3 statics :
Step 1= 0.0006 correct
sted 2= done 3 times to get good values= 0.020801. Indeed sometimes the values printed was very big.
Step3= first loop= abord itself or printed NAN, redo several times to get a good value: 0.0035
step 4a= 200 loops to get 0.103284
go back to step 2, 3, 4 to have a best result. Step 4a =0.091945 after 60 loops.

step 4b= 0.092082 after 400 loops and return to step 3 and step 4a,b to get 0.091934.
step 5= 0.091924 no problem
step 6= 0.0832233

Retry to get best values going bak to step 4 .