[startrek66]

I have welded the two jst connectors to gps / mag uBlox. I will connect the gates of the Revolution following the instructions in wiki (gps +I2c). I would then check later with gcs that the GPS device / mag functions. I do not care for now calibrate because they are within the home and know that you can not calibrate indoors. How can perform it?

[f5soh]

You can calibrate indoor with sometimes distorted results due to magnetic devices around but can tell you if i2C is working simply looking at Magnetometer tab / Auxmag alarm if alive.

[startrek66]

You can calibrate indoor with sometimes distorted results due to magnetic devices around but can tell you if i2C is working simply looking at Magnetometer tab / Auxmag alarm if alive.

I performed the calibration following your recommendations in the video and staying inside the house. The result I wanted to achieve was to be sure that Mag aux worked. I am attaching the picture and you wonder why even if the mag alarm of the external Mag exceeds 1% (variable 0.5-2%), the field remains evergreen?

[f5soh]

Default Warning / Errors levels are set to 5% and 15% so both Mags are green.

[Mateusz]

This is really strange mag orientation. Are you sure your AuxMag is oriented 40 pitch and 29 yaw ? Isn't that strange ? I would think it's always +-90/180/270 deg and rarely 45 deg on yaw for diagonally soldered Mag. But if your Mag is mounted flat with respect to FC, I find this numbers unjustifiable.

Roll means your AuxMag is up-side down, pitch indicates that it's 40 deg up (strange) and yaw 29 indicates it's rotated (is arrow on module pointing the same direction as FC arrow?) and both FC and AuxMag board are in the same plane ?

[TheOtherCliff]

I bet large money that those mag orientation numbers are wrong.  And that will NOT fly if it is that much wrong.  Numbers need to be zero (or just close) no matter what angle bank the FC has.

99% of people need to set them either:
- 0,180,0 (or 180,0,180 which is the same) if they have I2C mag (separate port connected to FlexiPort or I2CPort).
- 0,0,0 if they have either OP GPSv9 or DJI Naza GPS.

[startrek66]

This is really strange mag orientation. Are you sure your AuxMag is oriented 40 pitch and 29 yaw ? Isn't that strange ? I would think it's always +-90/180/270 deg and rarely 45 deg on yaw for diagonally soldered Mag. But if your Mag is mounted flat with respect to FC, I find this numbers unjustifiable.

Roll means your AuxMag is up-side down, pitch indicates that it's 40 deg up (strange) and yaw 29 indicates it's rotated (is arrow on module pointing the same direction as FC arrow?) and both FC and AuxMag board are in the same plane ?

You're right. I also found these strange values. the GPS module is mounted on the pedestal attached to the frame. I think the mistake was when calibrating the magnetometer. I have not always placed the frame perfectly perpendicular to floor. can this be the cause?

[startrek66]

This is really strange mag orientation. Are you sure your AuxMag is oriented 40 pitch and 29 yaw ? Isn't that strange ? I would think it's always +-90/180/270 deg and rarely 45 deg on yaw for diagonally soldered Mag. But if your Mag is mounted flat with respect to FC, I find this numbers unjustifiable.

Roll means your AuxMag is up-side down, pitch indicates that it's 40 deg up (strange) and yaw 29 indicates it's rotated (is arrow on module pointing the same direction as FC arrow?) and both FC and AuxMag board are in the same plane ?

You're right. I also found these strange values. the GPS module is mounted on the pedestal attached to the frame. I think the mistake was when calibrating the magnetometer. I have not always placed the frame perfectly perpendicular to floor. can this be the cause?

I redid the calibration and I checked that indeed the orientation of the arrow on the external mag is oriented with 90 degrees of difference (ref revolution). The external mag sensor is inverted by 180 degrees. therefore I entered the correct values. I also checked with the scope that the internal and external axes are synchronized. that I noticed is that in any case turning into space quadcopter the bars do not remain fixed at zero.

[Mateusz]

Scope look good and all axes look synchronized. To me values make sense, 180 roll flips the Mag, and  -90 rotates it in one direction (it's either 90 or -90).
Aux and internal mags are green and around 1% is good, also values so close to zero such as -2 indicate that mag is slightly rolled and -1 slightly yaw, while horizontal plane is aligned well. This is probably due to physical imperfections how board is glued inside the case or mounted etc.. I would not bother with deviations lower then +-2 and keep correct rotations as whole 90/180 numbers as this is what you verified by opening GPS module and looking on the chip. Small errors such that should be tolerable, but wrong orientation like you had before by more degrees wouldn't fly like Cliff mentioned in previous post. In short, looks good to me now.

[startrek66]

Scope look good and all axes look synchronized. To me values make sense, 180 roll flips the Mag, and  -90 rotates it in one direction (it's either 90 or -90).
Aux and internal mags are green and around 1% is good, also values so close to zero such as -2 indicate that mag is slightly rolled and -1 slightly yaw, while horizontal plane is aligned well. This is probably due to physical imperfections how board is glued inside the case or mounted etc.. I would not bother with deviations lower then +-2 and keep correct rotations as whole 90/180 numbers as this is what you verified by opening GPS module and looking on the chip. Small errors such that should be tolerable, but wrong orientation like you had before by more degrees wouldn't fly like Cliff mentioned in previous post. In short, looks good to me now.

Thanks for the comments. This calibration I made indoors at home I'll try to repeat it outside for a better result.

[startrek66]

Scope look good and all axes look synchronized. To me values make sense, 180 roll flips the Mag, and  -90 rotates it in one direction (it's either 90 or -90).
Aux and internal mags are green and around 1% is good, also values so close to zero such as -2 indicate that mag is slightly rolled and -1 slightly yaw, while horizontal plane is aligned well. This is probably due to physical imperfections how board is glued inside the case or mounted etc.. I would not bother with deviations lower then +-2 and keep correct rotations as whole 90/180 numbers as this is what you verified by opening GPS module and looking on the chip. Small errors such that should be tolerable, but wrong orientation like you had before by more degrees wouldn't fly like Cliff mentioned in previous post. In short, looks good to me now.

Thanks for the comments. This calibration I made indoors at home I'll try to repeat it outside for a better result.

Positive test flight with the position hold mode. Thank you for the result some experts such f5soh, theothercliff, Mateusz, hwh, what they gave me lots of advice for both the type of gps that the settings to make. Compass calibration was performed in the flight zone using the USB cable, but away from cars. The video shows the result. I had no problems due to magnetic fields and I made a neat conductor wiring. I still have any questions to the experts, but for now I am satisfied with the result. Thank you.

Compass status



Video

[drmeatball]

What model Revo are you using? You seem to have a good altitude hold :up:

[startrek66]

What model Revo are you using? You seem to have a good altitude hold :up:

I bought a Revolution. I do not know other details. what do you mean by the model?

[TheOtherCliff]

Where did you buy it so he can get one that he knows works that well. 

[startrek66]

Where did you buy it so he can get one that he knows works that well. 

Ok. This is the link

F16084 OpenPiolot CC3D Rivoluzione Controllore di Volo Revo Supportano la Funzione Auto-Ritorno FAI DA TE Mini 250 Quadcopter Multirotor Elicottero Ala Fissa
 http://s.aliexpress.com/F3eINZZ7