Comment by Michael Zaffuto on December 30, 2009 at 3:09pm

Hi Jordi,

I am using the HMC5843 sparkfun breakout. It had a 4.7uF tantalum capacitor in that location. The device was unable to work with that capacitor. After finding others on the SparkFun forums with the same problem, the solution was to replace the tantalum with a low-ESR ceramic capacitor. The size A tantalum had essentially 1206 sized footprint. I had two capacitors on hand, 10uF, 25V, X5R 1206 (digikey 587-1337-1-ND)
and a 100uF, 6.3V, X5R 1206 (digikey 490-4539-1-ND). I decided to use the 10uF capacitor, and once installed..everything worked!! Apparently there is a repetitive narrow spike of current required by the magnetometer so a good low-esr cap is required. My 10uF capacitor seems to be tall as the one shown your picture and it doesn't appear to be causing a problem.

I'm currently working on small battery powered rf link to be able to do better rotational testing.

3D Robotics Comment by Jordi Muñoz on December 30, 2009 at 3:16pm

Crap i just correct the tittle of the post.

Hello Michael,

Yes i had the same problem, i just soldered a huge electrolytic one parallel to the original 10uF capacitor and the problem was gone. In this case i will use this 100uF ceramic capacitor (just in case).

Maybe your are right because mainly the capacitor is "ceramic" material, so it should not block/reflect magnetic fields.

3D Robotics Comment by Jordi Muñoz on December 30, 2009 at 3:18pm

thanks for the help!! ;-)

Comment by Michael Zaffuto on December 30, 2009 at 3:26pm

I have been testing it in my basement, surprisingly repeatable and works well. I have serial output of X,Y and Z sensor readings as well as the total magnitude mag=sqrt(sq(x)+sq(y)+sq(z)). I do see an elliptical nature to the magnitude as can be expected and can be corrected (scale factors and offsets). This is why I need my rf link working, to freely rotate sensor for data.

Comment by Zoltan Beldi on December 30, 2009 at 7:07pm

The capacitor will only contribute a "hard" error by virtue of the leads that may have some small ferrous component.
Otherwise, there will be much more pervasive errors from nearby components that are likely to acquire permanent magnetic fields, not to mention the induced magnetic fields from nearby cables carrying up to 2 amps of current. (as in Electronic control units for the motors)
However, the actual precision and accuracy of the HMC5843 isn't all that good anyway, so those field variations would probably get lost in the noise inherent in the system.

Comment by Michael Zaffuto on December 30, 2009 at 8:12pm

I was going to go with one of these bad boys,
but the EasyStar cringed at the 1000grams, :)

G-822A CESIUM MAGNETOMETER

3D Robotics Comment by Jordi Muñoz on December 31, 2009 at 10:20am

Zoltan,

Thanks for the info!

3D Robotics Comment by Jordi Muñoz on December 31, 2009 at 10:21am

1Kg magnetometer, thats for an aircraft???

Comment by Morli_ on December 31, 2009 at 4:40pm

Hi Jordi , I can't help you with choice of the capacitors problem :( but I see the DIYDRONEs symbol is in the break away portion , may be you want to put it some where in the main board :-) , may be you have done it already , just in case.

3D Robotics Comment by Jordi Muñoz on December 31, 2009 at 8:16pm

Its also in the bottom of the board. ;-) Thank.

Comment

• ‹ Previous
• 1
• 2
• Next ›
• Page