Jacques Caron writes:
> Hi,
>
> At 04:40 15/01/2004, James Ewing wrote:
> >You can read the Broadcom specifications for the 2050 radio chips in the
> >Linksys WRT54G here -
> >
> ><http://www.sveasoft.com/postt48.html>http://www.sveasoft.com/postt48.html
> >
> >84 mw is well within their design limits.
>
> Read the PDF linked, not the (wrong) summary in the post which, as far as I
> understand it, gives power... consumption of the card!
>
> Also, check out the (real) specifications in the PDF. The 802.11a chip has
> an internal PA, and the Tx power limit is 18.7 dBm (74 mW) up to 36 Mbps,
> and goes down to 12.8 dBm (19 mW) at 54 Mbps! Given that higher rates
> require a better signal at the far end (Rx sensitivity gets worse with
> higher rates), you don't want to go the 54 Mbps route for long links!
This backing off of power is not uncommon with OFDM. PAPR issues mean
that either you run a PA with more linearity (expensive) or a higher
IP3 (also expensive, potentially power hungry) or your EVM goes to
h*ll at the higher order modulations. In addition, spectral regrowth
may cause OOB emissions at the band edges.
> Without full specs for the PA used, and more importantly, testing as shown
> in the PDF, you might want to limit your ambitions a little bit. Instead,
> consider separating rx and tx antennas (with high gain rx antennas) to stay
> within legal limits and get the longest range.
Un fortunately, most 11g chipsets barely meet OOB emission limits
(which, as I said, are strict power limits, not relative) with the
2.2dBi dipoles that come with the unit.
Bigger antennas (on tx, anyway) may well generate the same out of band
emissions that are the subject of this thread.
Jim
--
"Speed, it seems to me, provides the one genuinely modern pleasure."
-- Aldous Huxley (1894 - 1963)
--
general wireless list, a bawug thing <http://www.bawug.org/>
[un]subscribe: http://lists.bawug.org/mailman/listinfo/wireless
