I'm getting similar reading variations from the 2 DIY Drone HMC5843 REV1.1 boards I purchased in NOV 2010.
Below is the response from Chris Anderson followed by my e-mail to him after seeing his response to your posting here.
I haven't sent my e-mail to Jordi yet because I haven't figured out how to do it without having to retype the whole thing.
Perhaps I'll try sending it as an attached file Cut and paste isn't working .
I'll let you know if I learn anything. Likewise, if you have a breakthrough in obtaining more accurate or better repeatable readings from the HMC5843 please let me know.
James Harris
Chris Anderson has sent you a message on DIY Drones
Subject: +/- 5 degree variances in HMC5843 outputs
------------ I'm probably not the right person to ask (I didn't develop that board). Maybe ask Jordi?
I see this sort variance on the bench due to all sorts of magnetic noise issues, ranging from the connecting cable to other equipment on the bench. It doesn't seem to be a problem in the air once we run it through filters and static-iron calibration.
But again, I'm not the expert on this, so I suggest you ask Jordi. Have you tried any other mags?
> Date: Mon, 20 Dec 2010 19:12:39 +0000 > From: James Harris > To: Chris Anderson > > Hi Chris, > > You recently responded(very briefly) to another customer's message about HMC5843 +/- 5 degree reading variances. > I'm getting similar responses from the two DIY Drones HMC5843 REV 1.1 boards I purchased back in Nov. A 10 degree range in repeatability when the magnetometer is stationary, flat and level and not near metal or other offending objects AND not in a varying magnetic field would not seem to be what should be "expected". > The HMC5843 spec is oddly missing any reference to repeatability but other Honeywell docs on HMC5843 (it's marketing presentation I believe) says at least a 1 degree accurate compass is possible with HMC5843. With the reading variances I'm seeing, nothing close to 1 degree is likely. (other Honeywell docs suggest a magnetometer with 12-bit ADC's could resolve 0.1 degrees) > On one of my HMC5843 boards I am seeing over 100 raw counts variance in readings (in 10 to 20 consecutive readings) in each axis EVEN when the POSITIVE or NEGATIVE BIAS is applied (the other board's readings vary by about 50). (Again... with board stationary, flat, level and not in the Bermuda Triangle (i.e. not in varying magnetic field). > With these large variances, how in the world can we obtain VALID gains and offsets and get a reliable heading? ( Is 10 degrees (+/- 5 degrees) variance considered to be a reliable heading? ) > > Any suggestions would be appreciated. > > Thanks, > James > >
Replies
Now it change to this.
if i rotate it pass certain angle the 2nd data change to -4498
Again...
Are the raw magnetometer data normal?
the third one just getting 5239.
"Noise" I've easily removed by averaging 4 to 8 samples using a ring buffer.
Repeatable variance has to be calibrated for.
http://www.diydrones.com/forum/topics/automatic-nulling-of?xg_sourc...
Hi Nick,
I'm getting similar reading variations from the 2 DIY Drone HMC5843 REV1.1 boards I purchased in NOV 2010.
Below is the response from Chris Anderson followed by my e-mail to him after seeing his response to your posting here.
I haven't sent my e-mail to Jordi yet because I haven't figured out how to do it without having to retype the whole thing.
Perhaps I'll try sending it as an attached file Cut and paste isn't working .
I'll let you know if I learn anything. Likewise, if you have a breakthrough in obtaining more accurate or better repeatable readings from the HMC5843 please let me know.
James Harris
Chris Anderson has sent you a message on DIY Drones
Subject: +/- 5 degree variances in HMC5843 outputs
------------
I'm probably not the right person to ask (I didn't develop that board). Maybe ask Jordi?
I see this sort variance on the bench due to all sorts of magnetic noise issues, ranging
from the connecting cable to other equipment on the bench. It doesn't seem to be a
problem in the air once we run it through filters and static-iron calibration.
But again, I'm not the expert on this, so I suggest you ask Jordi. Have you tried any
other mags?
> Date: Mon, 20 Dec 2010 19:12:39 +0000
> From: James Harris
> To: Chris Anderson
>
> Hi Chris,
>
> You recently responded(very briefly) to another customer's message about HMC5843 +/-
5 degree reading variances.
> I'm getting similar responses from the two DIY Drones HMC5843 REV 1.1 boards I
purchased back in Nov. A 10 degree range in repeatability when the magnetometer is
stationary, flat and level and not near metal or other offending objects AND not in a
varying magnetic field would not seem to be what should be "expected".
> The HMC5843 spec is oddly missing any reference to repeatability but other
Honeywell docs on HMC5843 (it's marketing presentation I believe) says at least a 1
degree accurate compass is possible with HMC5843. With the reading variances I'm
seeing, nothing close to 1 degree is likely. (other Honeywell docs suggest a
magnetometer with 12-bit ADC's could resolve 0.1 degrees)
> On one of my HMC5843 boards I am seeing over 100 raw counts variance in readings
(in 10 to 20 consecutive readings) in each axis EVEN when the POSITIVE or NEGATIVE BIAS
is applied (the other board's readings vary by about 50). (Again... with board
stationary, flat, level and not in the Bermuda Triangle (i.e. not in varying magnetic
field).
> With these large variances, how in the world can we obtain VALID gains and offsets
and get a reliable heading? ( Is 10 degrees (+/- 5 degrees) variance considered to be a
reliable heading? )
>
> Any suggestions would be appreciated.
>
> Thanks,
> James
>
>
Yes.