--pgp-sign-Multipart_Mon_Nov__1_17:41:12_2004-1
Content-Type: text/plain; charset=US-ASCII

>>>>> "pc" == Pavel Cahyna <pavel.cahyna@st.cuni.cz> writes:

    pc> What is your experience?

    pc> How does the manufacturer himself comply with the regulations,
    pc> given the influences you named? I'm just curious.

Note that one of the things HAMs are always harping about is that the
FCC certifies complete systems only, so you're not technically allowed
to mix-and-match amps, cards, and antennas the way people do, unless
you're some kind of competent certified integrator with test
equipment.  What I mean is, the Orinoco high-gain antenna is certified
for use with Orinoco cards.  You're not allowed to just add antenna
gain and dBm or whatver and draw your own conclusions.

At a place I worked we installed commercial radio systems.  The power
output of the tranceivers varies from one serial number to another.
To maintain compliance/performance, we do two things:

 (1) attach a Bird calibrated RF power meter to the tranceiver output.
     Bird meters are very simple, but still expensive.  They're
     expensive because they're calibrated to units like mW rather than
     unitless.  We always perform this test and log the result.  It
     can help notice other results that are unreasonable, or equipment
     that drifts in performance over time.

 (2) put a patch antenna on the ground connected to a spectrum
     analyzer and look at the peak power at one frequency, or the area
     under the curve.  This is in practice unitless but can compare
     the eirp in one traffic lane to another.

This stuff is ~$10k per radio unit, and still any power output printed
in the manual is not meaningful.

so, I agree with David that anything printed in a manual is an
interesting guidepost but shouldn't be legitimized by configuring it
into a driver.  Indeed, the equipment I've seen (which isn't much
unfortunately) has only unitless power output settings.

Although, one device has settings in ``dB''.  the highest setting is
0.  so, it is still unitless, but you _can_ expect one setting being
half the power of another.  I dunno if the API should include that
idea or not.

-- 
Le fascisme est la dictature ouverte de la bourgeoisie.
		-- Georg Dimitrov

--pgp-sign-Multipart_Mon_Nov__1_17:41:12_2004-1
Content-Type: application/pgp-signature
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (NetBSD)

iQCVAwUAQYa7k4nCBbTaW/4dAQJ4qAQAladTr20t2/SvgKYA3Ebd0rjo8pDqHmQp
pbnHX3Cebjrwg+OgYDi5J+yKv1mHS+384XdZkrVUVr5sayN65b3hZh92F2L+KdLw
S1mX/pZg02ILcvdYBe4NinPO3cT6ZcJcu0NjqLLsz3cGfHSRxM6AL9xxcstDVLBC
4zMvxrr1G5s=
=yG3w
-----END PGP SIGNATURE-----

--pgp-sign-Multipart_Mon_Nov__1_17:41:12_2004-1--