on 5/13/02 1:52 PM, Paul Christensen at w9ac@arrl.net wrote:
> Looks like some great design work Jon. How about a quick history lesson on
> the product?
>
Paul and others,
See below for a response I gave another Amps member on the subject. His
question was if the module was really linear:
> Are they really liner???
Technically, no. The final module in the picture (the biggest one) is a
Class AB stage. The other two modules are EXTREMELY linear. The first
module is a 30 dB gain amplifier module that has an IMD rating of -75 dBc or
so. The second module we used to call the "pork chop." It is the flattest
PA device I've ever designed in terms of gain. It is flat to within less
than a tenth of a dB over the entire USA cellular band. I have a plot of it
somewhere!
The seller rated the module at 150 Watts. This is module is capable of that
as I believe the final transistor is an MRF899 or something like that. If I
remember correct, that's a 150 Watt class AB device. If memory serves me, I
think in the PA system, we ran this around 30 Watts or so - WELL backed off.
This module was a driver stage in Motorola's second generation Feed Forward
Linear Amplifier product line. Each PA system had 2 of these driver modules
running in a parallel/hot standby configuration. These then fed a total of
8 final PA modules. Each final PA module assembly consisted of 2 of the big
modules shown in the picture. In order to attain the linearity required by
the FCC, the whole thing was controlled by an error loop.
The error loop would sample the input signal, adjust gain and phase so as to
eliminate the wanted carrier signals that you do want in the main amplifier
path. So then in the error loop, you have just the intermod or distortion
products after removing the wanted carriers. These intermod products were
then amplified by the same amount as any intermod products in the main path.
Then they were shifted by 180 degrees in phase and added back to the main
path. And VIOLA! What happens when you add a signal of desired carriers and
undesired intermod with a signal with just the undesired intermod shifted by
180 degrees? You have just carriers left!
Pretty cool, huh?
The total PA system with all those final modules was theoretically capable
of at least 2400 Watts (each final transistor was capable of 150 Watts and
there were 16 of them total - 2 per each of the 8 final modules). However,
the system only ran at 150 to 160 Watts maximum power! So it was way backed
off. Intermods in the final output were down around -70 dBc or so so the
entire PA was VERY linear.
The amazing thing about technology now is that in the space that we had one
amplifier system delivering 150 Watts, I can now sell 3 amplifier systems
using the same technology that are capable of 600 Watts! And that module
that I designed is only about 7 years old!
There! How's that for an explanation? Make sense?
73,
Jon
NA9D
|