OK up until here Colin:
> "...The SWR reading gives no indication of reactive components, nor can it
> separate the resistive from the reactive components. It is calibrated
> with a pure resistive load and therefore has its greatest accuracy with
> pure resistive loads."
That's an old and very incorrect wive's tale. Incorrect rumor has it
that a 25 ohm load and 25 ohm R will look like 50 ohms and have a
1:1 false SWR reading, or that a 50 ohm impedance (which is not
the case with R25 j25) will somehow show as a 1:1 SWR when j is
not zero.
Whoever wrote that needs to spend some time on a test bench
looking at directional couplers.
A directional coupler responds very well to reactances. 50 ohms X,
or any value of X, can NEVER look like a 1:1 SWR in a directional
coupler that is nulled on a 50 ohm J zero load.
> "... It can be seen that a 3 to 1 SWR on the low side of 50 ohm will ask
> the amplifier to deliver much more power than a 3 to 1 SWR on the high
> side. Since the amplifier does have a finite value of output impedance,
> the amount of power delivered to the load with efficiency will change with
> load. Unless the load is near the design value, the transistors will heat
> up unnecessarily without delivering an more power to the antenna."
That's true. In one case you get excessive voltage, in the other
excessive current. Either one demands protection because either
one causes problems.
The excessive voltage problem causes flat-topping and splatter as
well as perhaps voltage breakdown in components, even though
dissipation in the PA transistors and power supply is less.
The excessive current problem generally has less of an impact on
linearity (until some device approaches current saturation) but
greatly increases dissipation and reduces efficiency.
Either way, it is foolish to allow a solid state PA to have a 3:1
mismatch at full power unless the PA is designed to optimum in
MORE than 50 ohms of load (perhaps 150 ohms) and has a lot of
headroom on current.
Then you'd suffer poorer efficiency at 50 ohms but the PA would be
acceptable over a wider range of mismatches.
73, Tom W8JI
w8ji@contesting.com
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|