That wasn't our experience.  The traces are symmetrical, and we used the
filter balun from Johanson.

We failed emissions by a large margin, and when we went looking for
problems, found that the antenna impedance was way off.  It was copied from
an app note, but that design was apparently for a PCB surrounded by lots of
air where ours has plastic within a few mms.  So we re-designed the antenna
for improved impedance match and the harmonic emissions are now 3dB or so
better than spec, and we are running wide open on TX power.  No filtering
other than the balun, which feeds the re tuned PIFA antenna.

We tried terminating with various impedances and verified that the
harmonics increased strongly if the return loss was worse than 12dB but
didn't have time to make measurements and plot a curve.




On Fri, May 17, 2013 at 2:40 PM, Joerg Wunsch <[email protected]> wrote:

> David VanHorn <[email protected]> wrote:
>
> > These chips also spew harmonics if the output impedance is too far
> > off.
>
> The key to a good supression of the second harmonics is to
> keep everything symmetric.  That way, most of the 2nd harmonics
> that still leave the internal PA will compensate outside.
>
> It's really recommendable though to add external lowpass filters.
> The baluns recommended in the datasheet do already have them.  But
> if you use a symmetric antenna, you better add two PI filters.
>
> (ETSI harmonics supression is just reachable even without them, but
> not much margin is left, so Tx power should be reduced a bit to gain
> margin.  FCC harmonics supression requirements are a bit harder, and
> always require a lowpass filter.)
>
> --
> cheers, Joerg               .-.-.   --... ...--   -.. .  DL8DTL
>
> http://www.sax.de/~joerg/
> Never trust an operating system you don't have sources for. ;-)
>
> _______________________________________________
> AVR-chat mailing list
> [email protected]
> https://lists.nongnu.org/mailman/listinfo/avr-chat
>
_______________________________________________
AVR-chat mailing list
[email protected]
https://lists.nongnu.org/mailman/listinfo/avr-chat

Reply via email to