digital compass modules

joerg joerg.twinklephone at
Wed Jan 23 13:25:33 CET 2008

Am Mi  23. Januar 2008 schrieb Nils Faerber:
> I doubt that such a device would work in the phone, sorry.
> The GTA01 contains three loadspeakers with magnets, AFAIK GTA02 will
> still contain at least two. Then there is massive EM radiation from the
> GSM antenna which will interfere and finally there is the Bluetooth
> module also emitting EM when used.
> Earth's magnetic field is very weak so almost *any* EM radiation near to
> the measuring device (the chip) will interfere.
GSM with no call established is intermittant if any RF (else the battery would 
drain quite quickly, in fact there's a burst every ~60min or so for T321 
refresh). BT may be switched off. The compass device may read out with 20Hz 
so you may allways get some good readings in between, even while established 
GSM connection. Furthermode RF is no static magnetic field, it shouldn't care 
at all. We are talking about "DC" here, not "HF".
For the speaker's magnets: they clearly are a strong source of error, but the 
datasheet says the chip might calibrate to compensate for this IIRC.
Anyway the chip should be positioned as far away from each such source of 
magnetic interference as possible. (for a first simple test, put a small old 
fashioned compass on top of the Neo and check it out. As long as you get any 
reproduceable orientation dependent steady reading, chances are good)
What you never may compensate is current generated fluctuating magnetic fields 
when e.g. GSM module is pulling random 1A from battery and the + and - leads 
aren't parallel/twisted or *magnetic* shielded. So compass readout probably 
should be done in low power mode of the NEO - no backlight, no GSM 
transmission, no full speed CPU.


More information about the community mailing list