From: email@example.com (Tom Bruhns)
Subject: Re: Should one worry about reflected current into untuned loads at low power?
Date: 4 Dec 2002 12:37:15 -0800
NNTP-Posting-Date: 4 Dec 2002 20:37:15 GMT
Paul Burridge wrote in message news:<firstname.lastname@example.org>...
> At what point in increasing power output from a PA does one need to
> worry about reflected power when squirting a RF signal down a
> radom-length wire? Is it possible to get away with say 6W out without
> damaging the output device?
"Reflected power" may be a poor way for you to think about this
problem. Instead, try this: whatever load you have on the output
terminals reflects back some load at the output active device (be it a
valve, FET, bipolar transistor, or whatever). That particular load
and whatever signal you're amplifying and the bias on the amplifier
will result in some particular power dissipation in the active device
(and in other circuit elements, too), and some particular power
delivered to the load. If you dissipate too much power in the active
device (or in some other component), it will fail prematurely. The
conditions will also result in a particular set of voltages and
currents, and if any of those is too great, it will cause trouble. In
a high-Q circuit, you can have some pretty big voltages and currents
with a pretty low power dissipation!
Normally, if the amplifier stage is designed for a particular load
impedance and your load is very much different from that, you'll be
well served to put a matching circuit (an antenna tuner!) between the
amplifier output and the load. How much different? It's probably not
a big deal for a 1.5:1 SWR (relative to the design impedance), and if
the PA is equipped with a tuneable tank, it may accomodate a fairly
large range of load impedances just fine.
As for a particular power level, how can we say? I can tell you for
sure that if the load is purely reactive, trying to deliver 6W of
actual power to it will be an exercise in futility. But a 1kW output
PA could probably deliver 6W to a horribly mismatched load without
problem, and a 1W PA would probably die quickly trying to deliver 6W
to even the best of loads. So "6W to the load" really isn't enough