Replies to This Discussion

Permalink Reply by Wouter van Verre on March 2, 2012 at 9:51am

Hi Mike,

Im not 100% suree, but Ithink that this problem could also be caused by a bad accelerometer...You said that you saw a constant roll, which might be caused by an incorrect signal from the accelerometer. The noise you describe might be the DCM algorithm trying to fight the bad accel.

Anyway, you say that you have another oilpan, it would be interesting to see what happens when you use that one.

Wouter

• ▶ Reply

Permalink Reply by Mike Goza on March 5, 2012 at 2:05pm

I swapped the oilpan with another and everything works fine.  I guess the original one was bad from the start.

 

Mike

 

• ▶ Reply

Permalink Reply by Dee Smith on March 5, 2012 at 2:30pm

HI Mike, 

are you using the BL_Ctrls for this set up, if so what did you have to change to get the code to work, I  still have the hexa_motors.ino you posted but when i compile it with old and new code I get a bunch of out of range errors and doesn't produce a sketch. also do you have other sensors in your setup ie mag optical flow, sonar ?.

• ▶ Reply

Permalink Reply by Mike Goza on March 5, 2012 at 2:51pm

Yes I'm using the BL_Ctrls V2.0 in this hexa.  I'm compiling with arduino 022-relax patch.  I'm running arducopter V2.4.  In order to get it to work in the latest version, I had to modify the inputs to the I2C motor setting function to match the new #defines, but that was all I did different.  I am using sonar.  I would like to use optical flow, but I'm probably going to roll my own.  My hexa is going to have a gumstix mounted on it with camera for image processing, optical flow and a few other things.

 

Mike

• ▶ Reply

Permalink Reply by Dee Smith on March 5, 2012 at 3:42pm

Mike, 

would you mind posting the changed files, I'm kinda new to this and so not very comfy with to know where I'm going wrong. Thanks

• ▶ Reply

Permalink Reply by Mike Goza on March 5, 2012 at 3:48pm

The only file I touch is the motors_hexa.pde.  I just changed it for V2.4.1.  I will test tomorrow, but I've done this a bunch of times with no issues.  This code assumes you are using the MK power board and the BL-CTRLs V2.0 have the addresses set as marked on the power board.  I have attached the file.

 

Mike 

Attachments:

• motors_hexa.pde, 8 KB

• ▶ Reply

Permalink Reply by Mike Goza on March 5, 2012 at 3:52pm

ooops!  Just noticed a small error.  Here is the corrected file.  Sorry about that.

Attachments:

• motors_hexa.pde, 8 KB

• ▶ Reply

Permalink Reply by Dee Smith on March 5, 2012 at 4:04pm

Its all good thanks

• ▶ Reply

Permalink Reply by Dee Smith on March 6, 2012 at 10:17am

Mike,

Are you using pullup resistors on the clock and bus lines or simply connecting your distribution board scl & sda to the APM clock and bus. If you are using resistors what are they rated, I haven't gotten my setup to control the BL_crtl yet and was thinking that may be the issue and was planing on adding some 10k resistors, or that the version 1.2 BL_Ctrls have a different address - I have been reading some documentation and blogs with the addresses at 0x52, 0x54, 0x56, 0x58, 0x5a, 0x5b. 

• ▶ Reply

Permalink Reply by Mike Goza on March 6, 2012 at 10:30am

I'm just connecting the scl & sda from the power board to the APM.  If you are using V1.2 BL_Ctrls, then the addresses are different and I have no clue as to what they are.  After lots of searching, I found the base address of the V2.0 which is what I am using.  You will have to search the net for that info.  The base address is important.  I just add the controller number (minus 1 I believe) to the base address to pick the motor I'm sending commands to.

 

Mike

Permalink Reply by Dee Smith on March 5, 2012 at 3:57pm

Awsome Mike, 

got it to compile with no errors, thanks for the help let me start my testing and see if everything works, I will let you know how it goes.

Dwayne

• ▶ Reply