How to calibrate the magnetometer of ADIS16480

Document created by analog-archivist Employee on Feb 23, 2016Last modified by analog-archivist Employee on Feb 23, 2016
Version 7Show Document
  • View in full screen mode

I need a more information about calibration of magnetometer of ADIS16480. How
to calibrate it? I don't understand your online document which is published on
the web "". If we use the excel sheet
"488-MagnSweeps-w-RegisterCalc.xlsx" for calibration, how to determine the user
input values (amplitude and offset of x and y, phase and noise values)?

I try to collect the testing data from ADIS16480 and put your calculation
spreadsheet. I can't reach a circle with my data. You can find the spreadsheet
on the attachment
("Z-Sistem_Magnetometer_HardandSoftIron_CalculationSpreadsheet.xlsx"). Also you
can find my readings on the other file ("Z-Sistem_Magnetometer_Values.xlsx"). I
want to check my readings. Please inform me; the values are correct or not? 
Will be new readings plot as a circle when I put the correction values of hard
and soft iron on it?

Firstly, I draw a disc which have 10degree part, as shown on attechment.
Printed out on A4 paper. After I atteched this paper on a flat wood part. I put
this wood part on a wood stool and I placed the testing platform on middle of
the room. Finally I return the aluminium casing on the paper with 10 degree and
collect the data from serial port. You can see the testing platform on the
attechment. I want to ask something about rotating. Must the center of the turn
be on the middle of the ADIS or PCB? and I collected the values of degree of
yaw, when turning the pcb, to check my process. It was compatible with turning

The other sensors works normally, becasue we are flying with that sensor. We
are using that sensor for stabilyzing of our UAV. We have 4 stroke engine and
either yaw, pitch, roll degree and barometer value are working correctly. We
need the magnetometer for auto take off. However only magnetometer is not
working good as you see. What is your opinion?

Must be the plane of the wood part, which using for calibration platform,
parallel on the ground?

We made some upgrade on our test platform and we collect new data. Could you
please investigate them. You can see on the attechment
If we have still some problems, I will read all register and send you.

I will read all registers today and send to you. I made a flat wood plate and I
mounted a battery, my power distribution board and IMU with own design PCB on
that plate. Only 3 cable going to outside from setup (rx, tx, gnd). After that
I turn the plate on a flat table, which is horizantal to ground and I read the
data with serial terminal, called "putty". The position of IMU lying flat on a
table and the connector pointing down. However I used the yaw degree output of
IMU for 10 degree turning.
I want to ask onething about magnetometer. I see a bad effect on magnetometer
output when connect a bluethoot tranciever to IMU board. The bluethoot
tranciever was placed 20cm far from IMU.  As you know the bluetooth devices has
low output power (approx. 10mW). Is it normal effect or not?
Also we have a 2W rf tranciever (2.4GHz) on our UAV and the antenna is placed
30cm far from IMU on the UAV. I think IMU will be affected by this antenna.
Could you provide some tips about avoiding the RF interference?

You can find all registers which were defined on datasheet, on the attachment
("IMU_ALL_REGS.xlsx"). X direction of IMU point to nearly North and Y direction
to South. I checked with a compass.


This was intended to be more instructional, not a "complete magnetometer
calibration solution".  We did post a second version of this, which offers more
insights on how to use the attached spreadsheets :

Your data looks like the device is not actually rotating.  How are you rotating
the part?

Can you provide a complete register dump and confirm that the other sensors
appear to be working normally? The device does not need to be in the center of
rotation for this experiment, because the stimulus is the earth's magnetic

I am wondering if you have loaded the magnetometer correction values,
covariance terms or magnetometer configuration registers with bad values.
Magnetometer failure is possible, but not most likely, based on what we have
seen so far.

In addition to the register dump, can you provide a picture of the label?

Yes, please keep the plane horizontal (for X-Y).

Some registers are more important than others, but it will be good to have a
complete picture of the device's status. Therefore please send me all the
register values you can read.

Pretty strange. What did you change in your setup? Looking forward to the
register dump - so please send all register values you can read. When you
capture these values, it would be good to know which direction the device is
pointed, along with confirmation of "device lying flat on a table, with the
connector pointing down."

I suspect that you will need to determine your antenna field patterns in order
to estimate the impact it can have on the fields, at the location of the
ADIS16480. At the low sample rates of the magnetometers, I suspect that
complete elimination of this impact will not be possible, but perhaps your
algorithms might be able to detect when they happen, so that you can reduce (or
turn off) the magnetometer influence when the transceiver is transmitting.

Please could you try with EKF_CFG = 0x0208 ?