Yaw bug / Magnetometer.

Hi.

I tried the FreeImu v 0.3 code for my Sparkfun Razor 9DOF .
But, i've got a problem with the yaw angle, wich is the outcome of strange magnetometer datas.
I film the situation : http://www.youtube.com/watch?v=Q2UiuMdofA0&feature=channel_video_title
As you can see, the board doesn't move, but the cube jiggle. ( I use quaternion code, and processing display).

I also tried the mongoose 9DOF firmware,but same results.

Do you have any ideas to help me? Thanks a lot.

Robin.

Never seen something like this. It may be an hardware issue. Would you mind try some of my gyro+accelerometers examples at http://www.varesano.net/blog/fabio/my-first-6-dof-imu-sensors-fusion-implementation-adxl345-itg3200-arduino-and-processing

I've got the same problem with the gyro/accel.

The orientation is globally good, but the cube jiggle.

But, I noticed that even if I unplug the 9DOF while the soft is running, ( no more datas incoming ), the cube still shaking.

I'm a bit lost.

Mmm.. me too. Have you modified the code somehow?

Would you mind trying using the MultiWii code? http://multiwii.googlecode.com/files/MultiWiiV1_8.zip
They have a nice Processing app which let you see the values coming from the sensors.

I'll try it right now, i'll keep you in touch.

Before trying the MultiWii code, I checked the test arduinos code for the 3 sensors. I noticed that both of Mag & gyro have problem.

First of all, the test program don't work with the 9DOF. The gyro.isRawDataReady() still 0 constant. So no gyro values.
Moreover, i tried the Raw HMC5883 soft. Here is the 10th first datas :
-15,-215,38,
27,-80,28,
120,-98,-69,
144,-189,4,
47,-169,14,
121,-21,-33,
92,-40,49,e
97,-53,71,
34,-57,3,
113,-197,16,

As you can see there is a serious trouble here. In my mind, the magnetometer values have to be similar, here the difference is quite huge.

I think my problems came from here, but dunno how to fix that.

 And here is a sample of Freeimu_Raw values that I receive :

-287,37, 26,5,34, 342,86,14,

1,-289,33, 25,7,33, 161,-44,-36,

-1,-285,35, 22,5, 3, 198,-91,-61,

4,-290,42, 26,5,33, 133,-335,-17,

0,-289,30, 19,1,29, 128,-365,-44,

-1,-290,24, 22,3,35, 248,78,-27,

-2,-286,27, 21,12,30, 127,133,-19,

1,-285,35, 22,9,35, 237,-112,-16,

2,-289,33, 26,8,35, 155,97,-39,

-2,-282,26, 21,4,31, 285,-87,-47,

1,-290,30, 18,1,35, 179,156,-23,

2,-286,38, 24,3,26, 476,54,-39,

Now I have gyro values. But still the magn doing strange things.

> As you can see there is a serious trouble here. In my mind, the
> magnetometer values have to be similar, here the difference is quite
> huge.

Were you keeping the IMU steady? If so, yeah, those data are
definitely strange.

But, this doesn't explain why you would see a shacking cube when the
IMU is not connected!

Are you next to some source of electromagnetic disturbances? Like a CRT screen, laptop speakers, halogen lights and stuff like that?

The Imu was of course steady. I though at the beginning it might be electromagnitc disturbance, but, the Imu is half a meter away from anything.

Ok, try with the MultiWii code.. if this also fails.. get in touch with Sparkfun guys and ask for replacement.

 I found somebody with the same issue.
http://forum.sparkfun.com/viewtopic.php?f=14&t=29350&p=131299&hilit=10736#p131299

So it's probably an hardware issue, maybe a solder, that's why I wrote an email to Sparkfun customer service to have their opinion on that point.
Let's see.

Ok, let me know how this goes.

Any update?

None.
So I took an ckdevice 9DOF fot the embedded arduino, and everything is fine.

Ok, I'm happy that you finally solved your question. You shuold however ask for a replacement board to Sparkfun.

Also, I wouldn't have supported the ckdevice.. they are spamming forums and blog about their products while their device aren't open.