1
\$\begingroup\$

I am currently working with a project where i am creating a small embedded board consisting of a 9DoF IMU which is transmitting its data with RF (BLE or normal RF). Furthermore there will be a battery attached extremely close to the unit.

I have few questions for this application:

  1. Can i have a RF antenna close to the IMU and how will it impact the electronic compass?
  2. Can i have a battery on the back or top of the board very close to it, and how will it impact the electronic compass?
    1. Which type of battery would be suitable: LiPo, Pure lithium, LiFe

I plan to use a Invensense IMU for the purpose.

\$\endgroup\$
1
  • \$\begingroup\$ What is the application? \$\endgroup\$
    – geometrikal
    Commented Mar 22, 2015 at 23:34

2 Answers 2

Reset to default
2
\$\begingroup\$

We sell a product that is very similar to what you propose.

  • RF should not affect it. We have a Bluetooth daughterboard to provide wireless comms and it does not affect the compass.

  • Temperature will affect the readings.

  • Metal parts such as USB connectors and SD card sockets will cause a magnetic distortion. Keep them away from the IMU.

  • Batteries will cause a large distortion depending on the size. We found Alkaline and Lithium AA batteries up to 10cm away can cause an effect.

  • If the batteries are close enough to have an effect, for high accuracy you will need to calibrate the magnetometer every time the batteries are changed. This is because different batteries, even of the same make and model, have different effects.

  • We found the calibration did not change as AA batteries discharged.

As for what battery to use, it depends on your power requirements. I think it is mainly the steel casing of the batteries that causes the biggest effect. Keep in mind that at very low power consumption alkaline batteries are just as good as lithium ones.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Very interesting input on this problem. I am glad to hear that RF won't have any impact on the signal. Currently i am aiming to use Lithium battery, but not 100% sure if it will be a simple coin-cell or a LiPo/LiFe. As far as i can understand, it should be enough to calibrate the compass with the battery attached and of course necessary to do so for every IMU (as you would normally do anyway). In other words i should be able to put it all together and calibrate it once, although i will have to re-calibrate in case the battery or any other part is replaced/changed. \$\endgroup\$
    – JavaCake
    Commented Mar 23, 2015 at 6:55
  • \$\begingroup\$ We also have a lithium coin cell (RTC backup) on the PCB and it can be a lot closer before having an effect, I think just a couple of cms. It is a bit of hack but you can can an idea of the effect using a mobile phone compass app and waving the battery around. :) But to answer your question, we have had instruments in the field for 3 months and magnetometer calibration does not appear to have changed. The sensor is only a few cams from the AA batteries too. If you can share, what is the application? \$\endgroup\$
    – geometrikal
    Commented Mar 23, 2015 at 11:18
1
\$\begingroup\$

Can i have a RF antenna close to the IMU and how will it impact the electronic compass?

The RF should have no significant effect on either the MEMS inertial sensors or the magnetometers.

Can i have a battery on the back or top of the board very close to it, and how will it impact the electronic compass?

The magnetic field associated with the DC current through the battery (or currents anywhere else on your PCB, for that matter) can introduce offsets into your magnetometer readings. Also, any ferromagnetic parts (e.g., iron, steel or nickel) of the battery, its holder, or any other nearby hardware can also introduce errors. Some of these errors can be calibrated out, so it's partly a question of how much effort you want to put into the software.

\$\endgroup\$
1
  • \$\begingroup\$ I have put significant effort into the software, but it has been a big questionmark for me in regards of the battery as i assume its not a constant offset that is introduced by the battery. In other words will the data be reliable even after the IMU has been calibrated with all the parts attached or will there be a significant tolerance from the offset over time while the battery is degrading or loosing its capacity? \$\endgroup\$
    – JavaCake
    Commented Mar 22, 2015 at 16:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.