werner at openmoko.org
Sat Jun 26 20:46:08 EDT 2010
JDH services wrote:
> It clearly says at he bottom of the ModIndex 1.0. MSK has a MonIndex
> of 0.5 in the real world.
I think you're right and what they implement isn't MSK. The
modulation index is indeed a good clue. What threw me off was
the tx deviation, which isn't the frequency shift one often
uses when talking about FSK, but the symmetrical deviation
from the FM carrier, i.e., half the frequency shift.
In fact, Silabs explain it quite well on page 28 of the data
Sorry for insisting this was MSK.
This configuration also has a name, it's Sunde's FSK ,
which helps with demodulation but isn't as spectrally
efficient as MSK. Nor is it compatible with OQPSK.
I haven't seen anything in the data sheet that says one
couldn't use a modulation index of 0.5, however AN466 
recommends on page 33 that the index be at least one.
 http://www.silabs.com/Support Documents/TechnicalDocs/AN466.pdf
More information about the discussion